G-CSF - Nature

Bone Marrow Transplantation (2001) 28, 259–264  2001 Nature Publishing Group All rights reserved 0268–3369/01 $15.00 www.nature.com/bmt

Haematopoietic growth factors Glycosylated vs non-glycosylated granulocyte colony-stimulating factor (G-CSF) – results of a prospective randomised monocentre study H Bo¨nig1, S Silbermann1, S Weller1, R Kirschke1, D Ko¨rholz2, G Janssen1, U Go¨bel1 and W Nu¨rnberger1 1

Department of Paediatric Haematology and Oncology, Center of Child Health, Heinrich-Heine University Medical Center, Du¨sseldorf; and 2Division of Paediatric Haematology and Oncology, Center of Child Health, University of Leipzig, Germany

Summary: The discovery of the haematopoietic growth factor granulocyte colony-stimulating factor (G-CSF) has reduced infection-related morbidity in cancer patients by alleviating post-chemotherapy neutropenia. Two formulations of recombinant human (rh) G-CSF, one glycosylated and one non-glycosylated, are available. The glycosylated form, lenograstim, possesses at least 25% greater bioactivity in vitro. Some comparative studies into the preparation’s potential to mobilise haematopoietic stem cells suggest a similar advantage. In the light of the great clinical importance of G-CSF, we have performed the first prospective, randomised, crossover study on children with chemotherapy-induced neutropenia. G-CSF (250 ␮g/m2) was started 1 day after the chemotherapy block, and was administered until a WBC ⬎1500/␮l was achieved on 3 successive days. Thirty-three G-CSF cycles from 11 patients (16 lenograstim, 17 filgrastim) were studied. They were investigated for duration of very severe (WBC ⬍500/␮l, 9 vs 9.5 days, lenograstim vs filgrastim, median) and severe leukopenia (WBC ⬍1000/␮l, 11 vs 11 days), infections (CRP ⬎5 mg/dl, 5 vs 5.5 days), infection-related hospital stay (11 vs 9 days) and antibiotic treatment (9 vs 9 days). Statistical evaluation by paired analysis could not detect any difference between treatment groups; the median difference for all end-points was zero. In summary, at least at 250 ␮g/m2, in terms of their clinical effect on neutropenia, the two G-CSF preparations appear to have identical activity. Bone Marrow Transplantation (2001) 28, 259–264. Keywords: G-CSF; glycosylation; neutropenia; cancer; pediatric

Human granulocyte colony-stimulating factor (G-CSF) is a naturally occurring growth factor which specifically targets Correspondence: Dr H Bo¨nig, Department of Paediatric Haematology and Oncology, Center of Child Health, Heinrich-Heine University Medical Center, Moorenstr 5, D-40225 Du¨sseldorf, Germany Received 27 March 2001; accepted 3 June 2001

haematopoietic cells committed to the neutrophil lineage (CFU-G).1,2 It also regulates some functions of the mature neutrophil, including chemotaxis, migration, and superoxide production.3 These effects are pivotal to the repeatedly demonstrated ability of exogenous G-CSF to reduce length and severity of post-chemotherapy neutropenia. In doing so, it could significantly reduce incidence and severity of neutropenic infection, thus decreasing morbidity and health care expenditure.4–6 G-CSF is also being used to mobilize haematopoietic stem cells into the peripheral blood in both healthy volunteers7,8 and chemotherapy-treated patients.9,10 The purification and molecular cloning of G-CSF were performed in the mid-1980s, and the first recombinant human (rh)G-CSF, filgrastim, was approved in the USA for clinical use in chemotherapy-treated cancer patients in 1991.11 Filgrastim is produced in E. coli. Unlike the natural G-CSF molecule, which is O-glycosylated at the Thr-133 position,12 it contains no sugar moiety, but is otherwise very similar in its amino acid sequence and steric conformation. The potency of filgrastim has been reported to be precisely the same as that of the natural G-CSF.13 Most clinical data on G-CSF has been gathered with filgrastim. Lenograstim was licensed in Europe and Japan in 1993.11 It is produced in Chinese hamster ovary cells, which has permitted the introduction of carbohydrate chains identical to those of the natural G-CSF molecule.12 Lenograstim is thus more similar to the natural G-CSF, but the clinical relevance of this remains to be determined. While it has been conclusively demonstrated that glycosylation is not vital for the biological activity of rhG-CSF,14 the carbohydrate moiety does appear to confer greater stability to the G-CSF molecule15– 17 by protecting the cysteine-17 sulfhydryl group.18 On the other hand, clearance of the glycosylated G-CSF molecule appears to increase with prolonged application.19 The effect of glycosylation of the G-CSF molecule on its potency in vivo is thus not predictable. The safety and efficacy of both preparations has been demonstrated independently,4–6,20,21 but comparative data are scarce. In vitro studies have suggested a more than 25% greater potency of lenograstim compared to filgrastim.14,22–24 This was believed to be the result of the inferior chemical stability of filgrastim, which resulted in a decreased half-life in culture media.14–16,22 This belief was supported by the experiments of Querol et al24 who demonstrated that daily

Glycosylated vs non-glycosylated G-CSF in neutropenia H Bo¨ nig et al

260

repletion of G-CSF in the media could abrogate the difference in potency. Some groups have suggested that lenograstim might also possess greater potency in vivo. Therefore, two comparative animal studies were performed. One study, performed by Nohynek et al25 at the laboratories of Rhoˆ ne-Poulenc Rorer, looked at the impact of the two G-CSF preparations on the recovery of cyclophosphamide-treated mice. In both the lenograstim and the filgrastim groups the leucocyte nadir was reached on day 4, a 1 day advantage over the control animals. On day 12, both groups had reached pre-treatment leucocyte counts. Another study investigated pharmacokinetics and maximum leucocyte counts in G-CSF-treated monkeys.26 With subcutaneous injection, bioavailability, plasma levels and maximum leucocyte counts were identical. These data not only failed to substantiate the proposition that lenograstim might be more potent than filgrastim, but also suggested that the decreased in vitro stability of filgrastim might not be observed in vivo. Comparative studies in humans have all looked at stem cell mobilisation, but their results were inconclusive. In a group of healthy volunteers, Høglund et al27 found lenograstim to be 25% superior to filgrastim in terms of CD34+ cell and CFU-GM mobilization. The same study design was used by Watts et al.28 However, in their study there was no difference in CD34+ cell mobilisation, but the number of GM-CFC was higher in the lenograstim-treated subjects. Watts et al28 also compared day-1 G-CSF plasma levels and found filgrastim levels to be significantly higher than lenograstim levels. These data again indicate that the decreased stability, which resulted in the inferior in vitro performance of filgrastim, should not play a role in vivo. However, comparison of plasma levels is not necessarily a good measure of effect, since plasma levels may only inadequately correlate with receptor binding. The authors therefore suggest other mechanisms involving ligand– receptor interaction by which a greater potency of G-CSF might be achieved. A number of studies have looked at stem cell mobilization in chemotherapy-treated patients, but failed to find a difference between the two preparations.29,30 Thus, while the in vitro situation is quite clear, there is some reason to doubt a difference in potency of lenograstim and filgrastim in vivo. Data comparing the two drugs’ potential for reducing neutropenia, morbidity from neutropenic infection, hospital days, and antibiotic treatment, is lacking altogether. In a prospective, randomised, crossover study, we have therefore compared blood counts, C-reactive protein values (CRP), and infection-related clinical events in patients receiving growth factor support with lenograstim or filgrastim following cytoreductive chemotherapy.

Materials and methods Patients Between August 1997 and August 1999, 11 patients, three female and eight male, aged between 3 and 29 years (median, 14 years) were included in the study. The investigation was performed as a monocentre study at the HeinBone Marrow Transplantation

rich-Heine University Center of Child Health, Department of Paediatric Haematology and Oncology. This prospective study was open for all patients who on clinical judgement were considered at ⭓50% risk of developing severe neutropenia, or were to receive rhG-CSF routinely as part of their chemotherapy protocol. Patient details are given in Table 1. Written informed consent was obtained from the patients or their guardians. The study was performed in accordance with the current version of the Declaration of Helsinki and had been approved by the local Ethics Committee under the condition that no additional blood draws or outpatient clinic visits should become necessary. Thirty-three cycles of rhG-CSF treatment were evaluated. Treatment RhG-CSF was started 1 day after the end of the chemotherapy block. It was administered once daily by subcutaneous injection at a standard paediatric dose of 250 ␮g/m2 ± 5%. A dose-response relationship for G-CSFsupported abbreviation of chemotherapy-induced neutropenia has been established for doses between 0.5 and 60 ␮g/kg,31,32 thus 250 ␮g/m2 (or 6 ␮g/kg) constitutes a medium dose. RhG-CSF was continued until a white blood count (WBC) of ⬎1500/␮l had been measured for 3 successive days after the end of the expected leucocyte nadir.

Table 1 Age, sex, diagnosis, and chemotherapy of the 11 patients included in this study No.

Age (y), Sex

1

28, M

2

29, M

3 4

20, M 7, M

5 6

14, F 12, F

7 8

3, M 22, M

9

18, F

10

8, M

11

10, M

Dx

Chemotherapy per cycle

VP16 1 × 2000, Melphalan 4 × 38 ARH Ifo 3 × 2000, VP16 3 × 150, VCR 1.2 ± ActoD 3 × 0.5 OS CarboPt 4 × 150, VP16 4 × 150 rERH CisPt 2 × 40, Ifo 4 × 1500, VP16 4 × 100, HT OS CarboPt 4 × 150, VP16 4 × 150 EWS Ifo 3 × 2000, VP16 3 × 150, VCR 1.2 ± (ActoD 3 × 0.5 or Adr 3 × 20) ERH Cyc 5 × 250, Topo 5 × 0.75 EWS Cyc 3 × 2000, VP16 4 × 55, tumor boost EWS Cyc 3 × 1800, VP16 4 × 100 ± tumor boost NBL DTIC 5 × 200, VCR 2 × 1.5, Ifo 5 × 1500, Adr 2 × 30 PNET CarboPt 4 × 200, VP16 4 × 100, (VP16 5 × 0.2 mg + AraC 1 × 30 mg i.th.) EWS

Cycles/pairs

2/1 3/2 3/2 2/1 2/1 5/4 3/2 3/2 4/3 2/1 4.3

Thirty-three cycles (22 cross-over pairs) of rhG-CSF therapy were investigated. All doses, unless otherwise specified, are mg/m2. Diagnoses: EWS = Ewing’s sarcoma; ARH = alveolar rhabdomyosarcoma; OS = osteosarcoma; (r)ERH = (recurrent) embryonal rhabdomyosarcoma; NBL = neuroblastoma; PNET = primitive neuro-ectodermal tumor. Chemotherapeutic agents: VP16 = etopside; Ifo = ifosfamide; VCR = vincristine; ActoD = actomycin D; Carbo/CisPt = carbo/cisplatin; HT = deep regional hyperthermia; Adr = adriamycin; Cyc = cyclophosphamide; Topo = topotecan; AraC = cytarabine.


Study design The trial was designed as a prospective, randomised, open, crossover monocentre phase-III study. The patients were treated alternatingly with lenograstim (Granocyte; Rhoˆ nePoulenc Rorer, Cologne, Germany) and filgrastim (Neupogen; Amgen, Munich, Germany). The randomisation for which preparation they would receive in the first cycle was made by lot. Two consecutive cycles with the same or very similar chemotherapies, but with different growth factor preparations, were to be investigated for differences by paired analysis. At every visit at the outpatient clinic, the patients were given a full physical examination, and a blood draw. For the evaluation, the following end-points or events were defined: leukopenia (WBC ⭐1000/␮l), severe leukopenia (WBC ⭐500/␮l), neutropenia (absolute neutrophil count (ANC) ⭐1000/␮l), severe neutropenia (ANC ⭐500/␮l), presumed infection (CRP ⬎5 mg/dl), infection-related hospital stay (hospital stay with antibiotic treatment; hospital stay for other reasons was disregarded), antibiotic treatment (treatment with antibiotics other than pneumocystis prophylaxis with co-trimoxazole). Elevated CRP rather than pyrexia was used as indicative of an infection, since a number of effects may influence body temperature, including ambient temperature and pyrexic (amphotericin B) or antipyrexic (metamizole, acetaminophen) effects of certain medications. The duration of rhG-CSF treatment was also monitored. If on some days no differentials were available, the ANC was nonetheless counted as ⬍500/␮l or ⬍1000/␮l, if the WBC was ⬍500/␮l or ⬍1000/␮l, respectively. For the evaluation of WBC (ANC) at one of the critical time points (500 or 1000 cells/␮l), no more than 2 days (3 days) were allowed to lie between two leucocyte counts (differentials), or the cycles would be excluded from the evaluation of WBC (ANC) kinetics. Methods Leucocyte counts were performed in the routine laboratory with an automatic cell counter. CRP was determined by standard laser nephelometry. If the WBC was ⬎500/␮l, blood smears were prepared and Pappenheim-stained, and differentials were hand-counted by light microscopy. Statistical analysis The condition for eligibility for crossover testing was that the same, or very similar chemotherapeutic agents, but different growth factors, had been given in two consecutive cycles. Two immediately consecutive cycles were compared in pairs using the Wilcoxon signed-rank-test for paired analysis. Differences were calculated as (time (days) with lenograstim treatment) − (time (days) with filgrastim treatment); thus negative ‘differences’ indicate a superiority of lenograstim, positive ones a superiority of filgrastim. In order to detect putative carry-over effects, the groups which had received, respectively, lenograstim or filgrastim first, were compared with the Mann–Whitney U-test for unpaired analysis. If more than one pair of cycles was gained from one patient, these were nonetheless considered independent

events. Thus, if a patient received drug A in cycle 1, drug B in cycle 2, and drug A in cycle 3, cycles 1 vs 2, and cycles 2 vs 3 were compared as two independent pairs. A P ⬍ 0.05 was considered significant, no adjustment for multiple testing was made. Confidence intervals and power analysis were calculated with the nQuery software. The study was estimated to be sufficiently powerful to detect as ‘statistically significant’ difference ⬎2 days for the parameters duration of WBC ⬍1000/␮l, WBC ⬍500/␮l, ANC ⬍1000/␮l and treatment duration (⬎3 days for ANC ⬍500/␮l) with a test significance level ␣ = 0.05, and a power of 80%.

261

Results Patient characteristics The 11 patients received a total of 33 cycles of G-CSF treatment (2–5 cycles, median three cycles per patient), 16 with lenograstim, 17 with filgrastim. This yielded 22 crossover pairs, of which 12 received lenograstim in the first, and filgrastim in the subsequent cycle, and 10 received filgrastim first, and lenograstim thereafter. In order to detect putative carry-over effects, the ‘lenograstim first’ and the ‘filgrastim first’ groups were compared with the Mann– Whitney U-test for unpaired analysis. No differences could be detected for any of the end-points. Thus, there appeared to be no carry-over effects, so that the order in which the growth factors were given did not have to be taken into account in the subsequent analyses. G-CSF was tolerated well; no major adverse effects were reported that were attributable to G-CSF treatment. For the paired analysis, in three patients minor differences between the chemotherapies were accepted. Thus, in one patient, the first of three cycles with ifosfamide, etoposide and vincristine was supplemented with actinomycin D, one patient was alternated between adriamycin and actinomycin D in addition to ifosfamide, etoposide, and vincristine, and in the third patient, the third and fourth cycles of high-dose cyclophosphamide and etoposide were supplemented with a radiation boost to the tumor bed (Table 1). Duration of leukopenia/neutropenia Sixty-nine percent (11/16) of the lenograstim, and 76% (13/17) of the filgrastim-treated patients dropped to a WBC ⬍1000/␮l, and in both groups on day 9 of growth factor treatment, ⭓50% of the patients (10/16 vs 10/17) had recovered a WBC of ⭓1000/␮l. Only 42%/41% (7/16 vs 7/17) of each of the treatment groups ever had a leucocyte nadir of ⬍500/␮l. In all patients, the ANC dropped to below 1000/␮l (n = 30 cycles); in both groups, day 11 was the first day on which ⭓50% of the patients (7/13 vs 10/17) had returned to an ANC ⭓1000/␮l. Very severe neutropenia (ANC ⬍500 ␮l; n = 29) was observed in 77% (10/13) vs 69% (11/16) of patients in each of the treatment groups, and on day 9, ⭓50% of patients had recovered their ANC to ⭓500/␮l in both groups (7/13 vs 8/16) (Figure 1). According to the power analysis, a difference between lenograstim and filgrastim for WBC ⬍1000/␮l and ANC Bone Marrow Transplantation


262 100

Lenograstim Filgrastim

90 80 70 60 50 40 30 20 10 0

a

Proportion of patients in the respective treatment group (%)


100


90 80 70 60 50 40 30 20

c

10 0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Time (days post chemotherapy)


100 90 80 70 60 50 40 30 20 10 0

b 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18



Time (days post chemotherapy) 100


90 80 70 60 50 40 30 20

d

10 0

1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18



Figure 1 Kinetics of WBC and ANC in chemotherapy-treated recipients of rhG-CSF. Eleven patients received 33 cycles of prophylactic rhG-CSF postchemotherapy (lenograstim, solid diamonds, or filgrastim, empty triangles) until a WBC ⬎1500/␮l had been observed for 3 successive days. Plotted to the X-axis is the time, to the Y-axis the percentage of patients with a WBC ⬍1000/␮l (a), WBC ⬍500/␮l (b), ANC ⬍1000/␮l (c), and ANC ⬍500/␮l (d). 44% of the lenograstim-treated and 41% of the filgrastim-treated patients had a WBC nadir ⬍500/␮l, 67 vs 76% had a WBC nadir ⬍1000/␮l, 77 vs 69% an ANC nadir ⬍500/␮l. All patients dropped to an ANC ⬍1000/␮l. In both groups, 50% achieved a WBC ⭓1000/␮l, ANC ⭓1000/␮l, or ANC ⬎500/␮l on days 9, 9, and 11, respectively.

⬍1000/␮l of ⬎2 days, equivalent to ⬎28% and ⬎18%, respectively, would have been detected. Comparison of pair differences yielded 0 (−2.3–1.3) days (median (95% confidence limits of the mean)) for the end-point WBC ⬎500/␮l, 0 (−3.3–0.9) days for WBC ⬎1000/␮l, 0 (−2.0–4.5) days for ANC ⬎500/␮l and 0 (−0.9–2.7) days for ANC ⬎1000/␮l. Box plots are displayed in Figure 2. Thus incidence, severity, and duration of leuko- and neutropenia followed the same pattern in lenograstim- or filgrastim treated-patients. Length of rhG-CSF treatment The duration of growth factor therapy was the same for both groups. Lenograstim was administered for 12.5 (7–18) days (median (range)), filgrastim for 12 (3–17) days. Power analysis revealed that a difference ⬎2 days for treatment duration would have been detected. This would have been equivalent to a difference of ⬎16% (test significance level ␣ = 0.05, power 80%). The difference among the pairs was 0 (−4–7) days (Figure 2). No difference was detectable between lenograstim- or filgrastim-treated patients. Infection Events defined as related to infection were observed in a minority of patients, but were equally common in both Bone Marrow Transplantation

treatment groups. Thus, infections (CRP ⬎5 mg/dl) were observed in 33% (5/15) vs 35% (6/17) of patients, they lasted 5 (4–9) (median (range)) vs 5.5 (3–12) days in lenograstim- vs filgrastim-treated patients. Median difference in the length of infectious episodes was 0 days, with a range of −8–7 days. The 95% confidence limits of the mean were −1.8–0.9. 58% (7/12) vs 60% (9/15) of patients received antibiotic treatment for an average of 9 (1–12) vs 9 (2–17) days (median (range)). The median difference was 0 days, with a range of −2–6 days. 25% (3/12) vs 50% (7/14) of patients were re-admitted or retained in hospital because of infections; the average infection-related hospital stay lasted 11 (6–12) vs 9 (3–14) days (Figure 2). Median (range) of the pair differences for infection-related hospital stay was 0 (−5–6) days, the confidence limits of the mean were −2.1– 1.2. None of these differences were statistically significant. Discussion For the in vitro system, a greater potency of lenograstim has repeatedly been demonstrated.14,17,22–24 This was, at least in part, explained by the greater pH resistance of the glycosylated molecule, since repletion of G-CSF could abolish the difference.17,24 A number of clinical studies have compared the two drugs’ potency for stem cell mobilisation, and they have


D t (lenograstim-filgrastim) (d)

20

10

0

-10

-20 n=

22 1

22 2

18 3

19 4

22 5

21 6

16 7

16 8

Variables Figure 2 Analysis of pair differences: consecutive chemotherapy cycles with different G-CSFs were analyzed as pairs (n = 22). Differences were calculated as ⌬ (days) = (duration with lenograstim (days) − duration with filgrastim (days)). Positive ⌬ thus indicate an advantage of filgrastim, and vice versa. Figure 2 shows box plots for the distribution of the ⌬. Differences are plotted to the Y-axis, the variables to the X-axis (1, WBC ⬍1000/␮l; 2, WBC ⬍500/␮l; 3, ANC ⬍1000/␮l; 4, ANC ⬍500/␮l; 5, duration of rhG-CSF treatment; 6, CRP ⬎5 mg/dl; 7, days with antibiotics; 8, infection-related hospital stay). The median ⌬ between lenograstim- and filgrastim-treated patients was 0 days for all end-points. Patients treated with lenograstim did not differ from those treated with filgrastim in terms of any of the end-points measured.

come to conflicting results. Thus in chemotherapy-treated, neutropenic hosts Schiødt et al29 and Saccardi et al30 found no difference in potency between the two preparations. However, these were non-randomised studies, and no power analysis was performed, so that it remains elusive how big the difference would have needed to be in order to be detected. In randomised cross-over studies with healthy individuals, lenograstim was reported to be more potent as a mobiliser of CD34+ cells27 and/or of GM-CFC.27,28 For all four studies, it must be said that the very subtle kinetics of G-CSF-induced stem cell mobilisation33 are probably not sufficiently reflected by once daily determination of CD34+ count or progenitor cell culture. The data available at present are thus not sufficient to justify any general statement about differences in potency of the two G-CSF preparations. An important negative feedback regulator of G-CSF is its neutralisation by mature neutrophils.34 This implies, that the pharmacokinetics of rhG-CSF will be quite different in neutropenic and non-neutropenic individuals.19,35 However, while the effect of the presence of neutrophils increases daily, the pharmacokinetic studies were only done on day 1 of G-CSF treatment.28 If glycosylated rhG-CSF was neutralised less effectively by neutrophils, this could explain better CD34+ or GM-CFC harvest and higher maximum WBC in probationers with very high neutrophil counts.25,27,28 During neutropenia, this difference would not come into play, which might explain why, even if there were a difference between the two GCSF preparations in terms of receptor binding, in neutro-

penic hosts consistently no difference was found between glycosylated and non-glycosylated G-CSF.29,30 As was mentioned above, the greater physical stability of the glycosylated molecule17 appears to confer to lenograstim an advantage in the in vitro system. In vivo, however, this appears not to be the case. Bioavailability with subcutaneous injection and plasma levels were identical in humans28 and monkeys.26 A number of factors may contribute to this. The stable physical environment and the continuous release from a subcutaneous depot may reduce the contribution of filgrastim’s lesser physicochemical stability. In vivo, clearance will much rather be the result of loss through proteolysis and neutralisation due to receptor binding and integration. The available studies suggest that these clearance mechanisms affect the two preparations similarly. Renal clearance of G-CSF has also got to be considered, since some of the cancer patients might have sustained relevant renal damage which would increase plasma levels.36 Differential effects of renal dysfunction on glycosylated and non-glycosylated G-CSF are conceivable, but were never studied. This is the first study looking at differential effects of the two preparations on the kinetics of chemotherapy-induced neutropenia in humans. We could here demonstrate that at a standard paediatric dose of 250 ␮g/m2, lenograstim and filgrastim did not differ in their effect on duration and severity of neutropenia, and on the frequency and clinical course of infections. The power of the study would have been sufficient to detect a difference in the range of that postulated by Høglund et al,27 ie a 25% advantage for lenograstim. Our data suggest that the in vitro dose equivalence of different rhG-CSF preparations may not be applicable in vivo, at least not in neutropenic patients. Rather, in children at 250 ␮g/m2, lenograstim and filgrastim are identical in their effects on neutrophil engraftment and the development of infectious complications. Differences in the biologic effects at other doses can not be ruled out. In children and adolescents, and at a dose of 250 ␮g/m2, the two drugs can probably be used interchangeably. This would permit the selection of the rhG-CSF preparation on the basis of vial size in relation to the patient’s body surface area, of cost, and of individual tolerability.

263

Acknowledgements We acknowledge the advice of Dr Willers, Computer Center of the Heinrich-Heine University, on the statistic evaluation of the data, and the excellent collaboration of the nursing, laboratory, and medical staff of the Department of Paediatric Hamatology and Oncology and the Laboratory for Haematology and Experimental Stem Cell Transplantation. This study was supported by Rhoˆ ne-Poulenc Rorer, distributors of lenograstim.

References 1 Bronchud MH, Potter MR, Morgenstern G et al. In vitro and in vivo analysis of the effects of recombinant human granulocyte colony-stimulating factor in patients. Br J Cancer 1988; 58: 64–69. Bone Marrow Transplantation


264

2 Lord BI, Bronchud MH, Owens S et al. The kinetics of human granulopoiesis following treatment with granulocyte colonystimulating factor in vivo. Proc Natl Acad Sci USA. 1989; 86: 9499–9503. 3 Yamamoto Y, Klein TW, Friedman H et al. Granulocyte colony-stimulating factor potentiates anti-Candida albicans growth inhibitory activity of polymorphonuclear cells. FEMS Immunol Med Microbiol 1993; 7: 15–22. 4 Crawford J, Ozer H, Stoller R et al. Reduction by granulocyte colony-stimulating factor of fever and neutropenia induced by chemotherapy in patients with small cell lung cancer. New Engl J Med 1991; 325: 164–170. 5 Trillet-Lenoir V, Green J, Manegold C et al. Recombinant granulocyte colony-stimulating factor reduces the infectious complications of cytotoxic chemotherapy. Eur J Cancer 1993; 29A: 319–324. 6 Welte K, Reiter A, Mempel K et al. A randomized phase-III study of the efficacy of granulocyte-colony stimulating factor in children with high-risk acute lymphoblastic leukemia. Blood 1996; 87: 3143–3150. 7 Schwinger W, Mache C, Urban C et al. Single dose of filgrastim (rhG-CSF) increases the number of hemopoietic progenitors in the peripheral blood of adult volunteers. Bone Marrow Transplant 1993; 22: 489–492. 8 Sato N, Sawada K, Takahashi TA et al. A time course study for optimal harvest of peripheral blood progenitor cells by granulocyte colony-stimulating factor in healthy volunteers. Exp Hematol 1994; 22: 973–978. 9 Duhrsen U, Villeval JL, Boyd J et al. Effects of recombinant human granulocyte colony-stimulating factor on hematopoietic progenitor cells in cancer patients. Blood 1988; 72: 2074–2081. 10 De Luca, Sheridan WP, Watson D et al. Prior chemotherapy does not prevent effective mobilisation by G-CSF of peripheral blood progenitor cells. Br J Cancer 1992; 66: 893–899. 11 Welte K, Gabrilove J, Bronchud MH et al. Filgrastim (rmetHuG-CSF): the first 10 years. Blood 1996; 88: 1907–1929. 12 Kubota N, Orita T, Hattori K et al. Structural characterization of natural and recombinant human granulocyte colony-stimulating factors. J Biochem 1990; 107: 486–492. 13 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. 14 Nissen C. Glycosylation of recombinant human granulocyte colony-stimulating factor: implications for stability and potency. Eur J Cancer 1994; 30A (Suppl. 3): 12–14. 15 Oh-Eda M, Hasegawa M, Hattori K et al. O-linked sugar chain of human granulocyte colony-stimulating factor protects it against polymerization and denaturation allowing it to retain its biological activity. J Biol Chem 1990; 265: 11432–11435. 16 Ono M. Physicochemical and biochemical characteristics of glycosylated recombinant human granulocyte colony-stimulating factor (lenograstim). Eur J Cancer 1994; 30A (Suppl. 3): 7–11. 17 Wang C, Eufemi M, Turano C, Giartosio A. Influence of the carbohydrate moiety on the stability of glycoproteins. Biochemistry 1996; 35: 7299–7307. 18 Arakawa T, Prestrelski SJ, Narhi LO et al. Cysteine 17 of recombinant human granulocyte colony-stimulating factor is partially solvent-exposed. J Protein Chem 1993; 12: 525–531. 19 Akizuki S, Mizorogi F, Inoue T et al. Pharmacokinetics and adverse events following 5-day repeated administration of lenograstim, a recombinant human granulocyte colony-stimulating factor, in healthy subjects. Bone Marrow Transplant 2000; 26: 939–946. 20 Gisselbrecht G, Prentice HG, Bacigalupo A et al. Placebocontrolled phase III trial of lenograstim in bone marrow transplantation. Lancet 1994; 343: 696–700.

Bone Marrow Transplantation

21 Stoppa AM, Blaise D, Viens P et al. Phase I study of in vivo lenograstim (rHuG-CSF) for stem cell collection demonstrates improved neutrophil recovery after autologous bone marrow transplantation. Bone Marrow Transplant 1994; 13: 541–547. 22 Pedrazzoli P, Gibelli N, Pavesi L et al. Effects of glycosylated and non-glycosylated G-CSFs, alone and in combination with other cytokines, on the growth of human progenitor cells. Anticancer Res 1996; 16: 1781–1785. 23 Mire-Sluis AR, Das RG, Thorpe R, participants of the collaborative study. The international standard for granulocyte colony-stimulating factor (G-CSF). Evaluation in an international collaborative study. J Immunol Meth 1995; 179: 117–126. 24 Querol S, Cancelas JA, Amat L et al. Effect of glycosylation of recombinant human granulocytic colony-stimulating factor on expansion cultures of umbilical cord blood CD34+ cells. Haematologica 1999; 84: 493–498. 25 Nohynek GJ, Plard JP, Wells MY, Zerial A. Comparison of the potency of glycosylated and nonglycosylated recombinant human granulocyte colony-stimulating factors in neutropenic and nonneutropenic CD rats. Cancer Chemother Pharmacol 1997; 39: 259–266. 26 Tanaka H, Tanaka Y, Shinagawa K et al. Three types of recombinant human granulocyte colony-stimulating factor have equivalent biological activities in monkeys. Cytokine 1997; 9: 360–369. 27 Høglund M, Smedmyr B, Bengtsson M et al. Mobilization of CD34+ cells by glycosylated and nonglycosylated G-CSF in healthy volunteers – a comparative study. Eur J Haematol 1997; 59: 177–183. 28 Watts MJ, Addison I, Long SG et al. Crossver study of the haematological effects and pharmacokinetics of glycosylated and non-glycosylated G-CSF in healthy volunteers. Br J Haematol 1997; 98: 474–479. 29 Schiødt I, Knudsen LM, Jensen L et al. Flow cytometry comparison of CD34+ subsets in bone marrow and peripheral blood after priming with glycosylated or non-glycosylated rhG-CSF. Bone Marrow Transplant 1998; 21: 1167–1170. 30 Saccardi R, Avanzi G, Bezzini R. Mobilization of PBSC for hematological rescue: comparison between glycosylated and non-glycosylated G-CSF. Bone Marrow Transplant 1997; 19: (Suppl. 11): 511. 31 Seymour AM, de Campos E, Thatcher N et al. A singleblind, randomised, vehicle-controlled dose-finding study of recombinant human granulocyte colony-stimulating factor (lenograstim) in patients undergoing chemotherapy for solid cancers and lymphoma. Eur J Cancer 1995; 31A: 2157–2163. 32 Neidhart J, Mangalik A, Kohler W et al. Granulocyte colonystimulating factor stimulates recovery of granulocytes in patients receiving dose-intensive chemotherapy without bone marrow transplantation. J Clin Oncol 1989; 7: 1685–1692. 33 Lee S, Im SA, Yoo ES et al. Mobilization kinetics of CD34+ cells in association with modulation of CD44 and CD31 expression during continuous intravenous administration of GCSF in normal donors. Stem Cells 2000; 18: 281–286. 34 Terashi K, Oka M, Ohdo S et al. Close association between clearance of recombinant human granulocyte colony-stimulating factor (G-CSF) and G-CSF receptor on neutrophils in cancer patients. Antimicrob Agents Chemother 1999; 43: 21–24. 35 Morstyn G, Foote M, Perkins D, Vincent M. The clinical utility of granulocyte colony-stimulating factor: early achievements and future promise. Stem Cells 1994; 12 (Suppl. 1): 213–228. 36 Tanaka H, Tokiwa T. Influence of renal and hepatic failure on the pharmacokinetics of recombinant human granulocyte colony-stimulating factor (KRN8601) in the rat. Cancer Res 1990; 50: 6615–6619.