Bone Marrow Transplantation (2003) 31, 371–378 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00 www.nature.com/bmt Stem cell mobilisation Successful mobilization of peripheral blood stem cells after addition of ancestim (stem cell factor) in patients who had failed a prior mobilization with filgrastim (granulocyte colony-stimulating factor) alone or with chemotherapy plus filgrastim

LB To1, J Bashford2, S Durrant3, J MacMillan4, AP Schwarer5, HM Prince6, J Gibson7, I Lewis1, B Swart1, J Marty4, T Rawling1, L Ashman1, S Charles4 and B Cohen4

1Division of Haematology, Hanson Centre for Cancer Research, IMVS, Adelaide, Australia; 2Wesley Medical Centre, Auchenflower, Queensland, Australia; 3Royal Brisbane Hospital, Herston, Queensland, Australia; 4Amgen Australia P/L, North Ryde, NewSouth Wales, Australia; 5BMT Program, Alfred Hospital, Prahan, Victoria, Australia; 6Department of Haematology, Peter MacCallum Cancer Institute, Melbourne, Victoria, Australia; and 7Institute of Haematology, Royal Prince Alfred Hospital, Camperdown, New South Wales, Australia

Summary: Bone MarrowTransplantation (2003) 31, 371–378. doi:10.1038/sj.bmt.1703860 This study assessed the ability of recombinant human stem Keywords: stem cell factor; stem cell mobilization cell factor (rHuSCF) to mobilize stem cells in 44 patients who had failed a prior mobilization (CD34+ yield 0.5– 1.9 Â 106/kg BW) with filgrastim-alone or chemotherapy- plus-filgrastim. The same mobilization regimen was used Autologous stem cell transplantation has improved the with the addition of rHuSCF. In the filgrastim-alone clinical outcome of patients with chemosensitive, inter- group (n ¼ 13), rHuSCF 20 lg/kg was started 3 days mediate-grade non-Hodgkin’s lymphoma or multiple mye- before filgrastim and continued for the duration of loma. Mobilization of peripheral blood stem cells with filgrastim. In the chemotherapy-plus-filgrastim group filgrastim (r-metHuG-CSF) alone or with chemotherapy (n ¼ 31), rHuSCF 20 lg/kg/day plus filgrastim 5–10 lg/ has been a major advance in improving the safety of kg/day were administered concurrently. Leukaphereses transplantation by increasing the number of harvested were continued to a maximum of four procedures or a hemopoietic stem cells. Despite this advance, some patients target of X3 Â 106 CD34+ cells/kg. In both groups, still fail to be mobilized and sufficient stem cells cannot be 1–4 CD34+ yield ( Â 106/kg BW) of the study mobilization collected for transplantation. was higher than that of the prior mobilization (median: Endogenous stem cell factor (SCF) is the natural ligand 2.42 vs 0.84 P ¼ 0.002 and 1.64 vs 0.99 P ¼ 0.001, for the proto-oncogene c-kit, a cell surface tyrosine kinase o 5 respectively). In all 54and 45%of patients in the receptor, and is an early-acting synergistic cytokine. filgrastim-alone group and chemotherapy-plus-filgrastim Recombinant human stem cell factor (rHuSCF) is a 166- group, respectively, reached the threshold yield of 2 Â 106/ amino-acid nonglycosylated protein. By itself, rHuSCF has kg. The probability of a successful mobilization was the no direct mobilizing effect, but when combined with same in those with a CD34+ yield of 0.5–0.75 Â 106/kg filgrastim and/or chemotherapy in nonrandomized studies, BW in the prior mobilization as in those with 0.76– mobilization was enhanced by three to five-fold in both 6–11 1.99 Â 106/kg BW. Downmodulation of c-kit expression untreated and previously treated patients. However, and a lower percentage of Thy-1 positivity in the enhancing the yield in patients who mobilized well does not mobilized CD34+ cells were noted in the successful guarantee that rHuSCF could mobilize in failed mobilizers. mobilizers compared with those in the poor mobilizers. We have previously shown that remobilization using the This study shows that rhuSCF is effective in approxi- same mobilization stimulus within 3 months produced mately half the patients who had failed a prior mobiliza- progenitor yields not different from the initial mobiliza- 12 tion and allows them to proceed to transplant. It also tion. While it is not ethical to repeat mobilization points to the likely role of the SCF/c-kit ligand pair in procedures in patients who have sufficient stem cells for mobilization. transplantation, repeat mobilization studies may be done in patients who fail to mobilize sufficient cells for transplanta- tion.13 This design would avoid the confounding effect of interpatient differences and tests rHuSCF’s efficacy by Correspondence: Dr LB To, Department of Haematology, Hanson Centre for Cancer Research, Institute of Medical and Veterinary Science, comparing the mobilization yield with and without Frome Road, Adelaide 5000, Australia rHuSCF in the same patient. Using a prior mobilization Received 17 July 2002; accepted 08 October 2002 in the same patient as a historical control should provide a Stem cell factor for failed mobilization LB To et al 372 powerful way to study the efficacy of mobilization regi- possible for this study, as patients had already failed prior mens. mobilization on filgrastim and/or chemotherapy alone In this multicenter study, we were interested to know if thereby eliminating the possibility of a comparative arm. rHuSCF can enable the collection of sufficient stem cells for As such, the study was historically controlled by comparing transplantation in patients who have failed conventional patient’s study yields with those from their previously failed filgrastim-alone or chemotherapy-plus-filgrastim mobiliza- mobilization. tion. Secondly, we investigated whether the downregulation Patients in the filgrastim-alone group received rHuSCF of c-kit is involved in successful mobilization in these failed 20 mg/kg/day administered for 3 days before the commence- mobilizers, a phenomenon seen after mobilization with ment of filgrastim and continued daily until leukaphereses filgrastim-alone or with chemotherapy-plus-filgrastim. were completed. A maximum of 11 days of rHuSCF administration was allowed. Filgrastim 10 mg/kg/day was administered for 5–8 days with leukapheresis commencing Patients and methods on day 5 of filgrastim administration. Patients in the chemotherapy-plus-filgrastim group received rHuSCF Patients and study design 20 mg/kg/day plus filgrastim 5–10 mg/kg/day 24 h after the The protocol was approved by the institutional ethics last day of mobilization chemotherapy, administered daily committees of the participating medical centers and all until leukaphereses were completed. Leukapheresis com- patients gave written informed consent before any study- menced when the white blood cell count was 42 Â 109/l or related procedures were done. Patients were eligible for the peripheral blood CD34+ count was X5 Â 106/l after the study, if they had nonmyeloid malignancies suitable for nadir. Four leukaphereses were scheduled for the study treatment with high-dose therapy, were at least 18 years of mobilization because previous studies have shown a more age, had an absolute neutrophil count (ANC) between 1.5 prolonged release of peripheral blood progenitors by and 10 Â 109/l, and Eastern Cooperative Oncology Group rHuSCF and filgrastim.14 A comparison of the study status of p2. Patients were required to have X0.5 but regimen with the previously failed mobilization might o2 Â 106 CD34+ cells/kg collected within the past 3 suggest a protocol requiring that patients undergo the months by at least three leukaphereses procedures after identical number of leukaphereses that were administered mobilization with a standard mobilization regimen of either under failed mobilization. However, leukapheresis yields filgrastim-alone or chemotherapy-plus-filgrastim. Whether from rHu-met-SCF mobilization have a tendency to the patient had filgrastim-alone or chemotherapy-plus- maintain yield levels across leukapheresis as opposed to filgrastim in this prior mobilization was determined by the filgrastim-mobilized yields, which are more likely to decline doctor treating the patient. However, patients received the progressively (data not shown). Hence, the study purposely same mobilization regimen plus rHuSCF during the study compares a previously failed study yield on at least three mobilization. Patients were excluded from enrolling in the leukaphereses with the protocol specification of a max- study, if they had any history of seasonal or recurrent imum of four leukaphereses. This avoids biasing against the asthma within the preceding 10 years; any anaphylactic/ potential of rHuSCF to produce meaningful yields in anaphylactoid-type event manifested by disseminated successive leukaphereses beyond what was achieved with urticaria, laryngeal edema, and/or bronchospasm; or any filgrastim alone. The study design is shown in Figure 1. history of angioedema or recurrent urticaria within the past Leukaphereses were performed on a Fenwal CS3000 plus 3 years. The study was performed between July and (Baxter, Deerfield, IL, USA) as described before1,12 and November 1999. were continued until a maximum of four procedures were performed or until a target of X3 Â 106 CD34+ cells/kg was reached. Separate pilot tubes were cryopreserved for Rationale for study design laboratory studies. In this study patients had rHu-met-SCF added to their Patients were categorized into two groups by the amount prior mobilization regimen. An active control was not of prior chemotherapy and/or radiation therapy received.

*Previous filgrastim 5-10µg/kg/day mobilisation with plus chemotherapy and filgrastim ancestim 20 µg/kg/day 5-10 µg/kg/day commencing 24 hours plus chemotherapy after chemo until completion of leukapheresis** OR Screening Enrolment

*Previous ancestim 20 µg/kg/day mobilisation with commencing 3 days filgrastim prior to filgrastim 10 10 µg/kg/day pg/kg/day until completion of leukapheresis##

Figure 1 Study schema.

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 373 Patients were classified as having had either extensive clinically relevant end point4 to assess the influence of therapy (42 cycles of procarbazine HCI, mechlorethamine potential prognostic indicators and the ability to proceed to HCI, nitrosourea, including BCNU, melphalan; high-dose transplant. cytosine arabinoside totalling X7g; X10 cycles of che- Wilcoxon rank sum test or permutation tests were used motherapy, or radiation to the mediastinum, abdomen, or for univariate comparison of continuous variables, and w2 pelvis, excluding spot radiation), or moderate/none (all or Fisher’s exact tests were used for univariate comparisons other patients). for categorical variables. For prognostic analysis, no more than a single variable was identified as associated with a 6 + Study drugs threshold yield of 42.0 Â 10 CD34 yield (sufficient yield for transplant); hence, multivariate analysis was not The rHuSCF used in this trial was produced using E. coli required. genetically engineered to express a methionine-terminated Patients who did not undergo leukapheresis were recombinant version of human SCF (ancestim; Stemgens, retained in the analysis and assigned a CD34+ yield of Amgen Inc., Thousand Oaks, CA, USA). Filgrastim zero. Patients who had more than the scheduled number of (Neupogens, Amgen Inc., Thousand Oaks, CA, USA) is four leukaphereses were included in the analysis with only also an E. coli-genetically engineered recombinant protein. yields from the first four leukaphereses. Patients receiving rHuSCF received premedication with ranitidine, bismuth citrate, salbutamol sulfate, and cetir- izine HCl starting 24 h before the first injection of rHuSCF Results and continuing every day of rHuSCF administration.10 Patients CD34+ and subset measurements Overall, 44 patients were enrolled (Table 1). Of these, 13 In the filgrastim-alone group, peripheral CD34+ counts subjects were previously mobilized with filgrastim-alone were performed daily until completion of leukaphereses. In and 31 with chemotherapy-plus-filgrastim. The filgrastim- the chemotherapy-plus-filgrastim group, peripheral CD34+ alone group contained more extensively pretreated patients counts were performed daily when the white blood cell (Po0.01, Fisher’s exact test) and the patients had a count was 41 Â 109/l and continued until the completion of marginally higher baseline platelet count (P ¼ 0.049, leukaphereses. Wilcoxon) than the chemotherapy-plus-filgrastim group. The two-color, immunofluorescence-staining procedure More patients in the chemotherapy-plus-filgrastim group and flow cytometric analysis have been previously de- had intermediate-grade non-Hodgkin’s lymphoma com- scribed.15 CD34+ cells were stained using HPCA2-PE pared with the filgrastim-only group, but fewer patients in (CD34) (Becton-Dickinson, Mountain View, CA, USA) the chemotherapy-plus-filgrastim group had a low-grade and were enumerated by backgating for CD45 expression non-Hodgkin’s lymphoma compared with patients in the (KD56-FITC, Coulter, Miami, FL, USA) and low side filgrastim-only group, but the differences did not reach scatter: CD34+ cell counts were performed on an aliquot of statistical significance. The CD34+ yields from the prior each day’s peripheral blood harvest. CD34+ cell yield was mobilization were similar in both groups. calculated based on the patient’s actual weight. The All 44 enrolled patients underwent mobilization. One coexpression of the following antigens was studied on patient in each group underwent five leukaphereses and one thawed cryopreserved cells on selected patients: stage- patient mobilized with chemotherapy-plus-filgrastim un- specific markers CD38 (Leu-17, Becton-Dickinson, Moun- derwent six leukaphereses as determined by the doctor tain View, CA, USA) and CD90 (Thy-1, PharMingen, San treating the patient, but the extra leukaphereses did not Diego, CA, USA); KDR (Sigma, St Louis, MO, USA),16 help to achieve the threshold yield of 2 Â 106 CD34+cells/ multiple drug resistance, MDR-1 (Immunotech, Prague, kg and were not included in the analysis. Two patients in Czech Republic), and stem cell factor receptor, CD117 the chemotherapy-plus-filgrastim group did not undergo (c-kit, PharMingen). The daily counts were performed in leukapheresis: one was withdrawn because of a moderate individual centers using similar methodology. The coex- allergic reaction; the other because peripheral blood pression studies were performed in one central laboratory CD34+ cell levels were insufficient to proceed to leuka- (Adelaide) over an 8-week period with the same batch of pheresis. reagents and the same test calibrations. Downregulation In the filgrastim-alone group, median CD34+ yield was measured by a comparison of the peak fluorescence of ( Â 106/kg) from the study mobilization was significantly the cell populations studied. higher than that of the prior mobilization (median 2.42, range 0.37–4.26 vs median 0.84, range 0.56–1.04, Po0.002), 6 Statistical methods and 7/13 (54%) reached the threshold dose of 2 Â 10 /kg. In the chemotherapy-plus-filgrastim group, CD34+ yield from CD34+ cell yields from rHuSCF-supported mobilizations the study mobilization was also significantly higher than were compared with prior CD34+ mobilization yield using that of the prior mobilization (median 1.64, range 0.0–5.52 the exact distribution for permutation tests for paired vs median 0.99, range 0.50–1.96, Po0.001) (Figure 2), and samples (StatXact, Cytel Software, Cambridge, MA, USA). 14/31 (45%) reached the threshold dose of 2 Â 106/kg. A threshold of 2.0 Â 106 CD34+ total (prior mobilization When the yield from the first mobilization (without plus study mobilization) cell yield was taken as the most rHuSCF) was added to the second (with rHuSCF), 85% of

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 374 Table 1 Patient demographics and baseline characteristics

Chemotherapy-plus-filgrastim Filgrastim-only (n=31) (n=13) Age (years) Median (range) 54 (19, 70) 55 (25,67) Sex Women 18 8 Men 13 5 Ethnic group (n/%) Aborigine 1 (3%) 0 Asian 1 (3%) 1 (8%) Caucasian 26(84%) 12 (92%) Other 3 (10%) 0 Disease type (n/%) Intermediate-grade non-Hodgkin’s lymphoma 12 (39%) 2 (15%) Hodgkin’s disease 3 (10%) 0 Low-grade non-Hodgkin’s lymphoma 5 (16%) 4 (31%) Multiple myeloma 8 (26%) 3 (23%) Other 3 (10%) 4 (31%) Prior therapy classification (n/%)* Extensive 4 (13%) 7 (54%) Moderate/none 27 (83%) 6(46%) CD34+ yield in prior mobilization (106/kg) Median (range) 0.99 (0.5–1.96) 0.84 (0.6–1.04) Baseline blood cell counts **Platelets (109/l) (median/range) 163 (86–347) 235 (143–344) Neutrophils (109/l) (median/range) 2.6(1–21) 2.7 (2–5) White cells (109/l) (median/range) 3.6(2–22) 4.1 (3–7) Mobilization chemotherapy CHOP 2 — DHAP 2 Cyclophosphamide 27 (2–4 g/m2)

**P=0.049; *Po0.01.

Prior Mobilisation patients were mobilized successfully with the addition of 6 With Ancestim rHuSCF. The percentage success rate is no different from those in the higher end of the range (0.76–1.99 Â 106 5 CD34+/kg). Hence, for a minimum prior mobilization yield of 0.5 Â 106 CD34+ cells/kg, no clear threshold 4 existed where a second rHuSCF-supported collection /kg) 6 would not be successful. 3 A total of 23 patients underwent transplant, 16in the chemotherapy-plus-filgrastim group and seven in the 2 filgrastim-alone group. Other 10 patients went into remis- CD34+(x10 sion and were not transplanted. These patients were still 1 receiving chemotherapy at the time of writing. Of the transplanted patients, the median (range) time to recovery 0 to 1.0 Â 109 neutrophils/l was 11 (4–19) days (n ¼ 21) and Chemotherapy (N=31) Filgrastim (N=13) the median (range) time to recovery to 20 Â 109 platelets/l (p<0.001) (p<0.002) was 11 (7–120) days (n ¼ 20). Figure 2 A comparison of the CD34+ yields between prior mobilization and the study mobilization in the two groups of patients. The bars show the 95th and fifth percentile, the box shows the 75th and 25th percentile and the Prognostic factors line shows the median. CD34+ cells/kg (2.0 Â 106) was used as the threshold to define successful mobilization.4 Using this threshold, the patients in the filgrastim-alone group had X2 Â 106 list of potential prognostic factors included prior mobiliza- CD34+/kg and 74% of patients in the chemotherapy- tion CD34+ yield, disease type, extent of previous plus-filgrastim group reached the threshold. If a ‘per- chemotherapy, cycles of radiotherapy, baseline hematology protocol’ approach, that is, including only the yield from (platelet, neutrophil and white blood cell counts), sex, and the first four leukaphereses, is used, 77 and 71% of the age. Only the baseline platelet counts appeared different filgrastim-alone and chemotherapy-plus-filgrastim patients, between patients who were successful and unsuccessful respectively, reached the threshold. mobilizers, with medians of 200 Â 109/l and 165 Â 109/l, In the subgroup of 14 patients whose prior yields were at respectively (P ¼ 0.05). In the chemotherapy-plus-filgrastim the lower end of range (0.5–0.75 Â 106 CD34+/kg), seven group, 12/21 patients (57%) whose baseline platelet counts

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 375 Table 2 A comparison of the phenotypes of CD34+ cells harvested in the first and the second mobilizations with patients segregated according to the success of the second mobilization

% CD34 CD34+ cell (106/kg) % CD38À % Thy-1+ % c-kit+ % MDR+ % KDR+ Poor mobilizers (n=8) Prior mobilization 0.15 (0.06–0.60) 0.33 (0.22–0.45) 8.05 (1.7–49.5) 15.45 (10.2–66.8) 34.3 (3.7–57.3) 2.85 (0.0–27.1) 0.55 (0.0–10.8) Study mobilization 0.33 (0.05–0.53) 0.34 (0.07–0.56) 4.60 (0.1–28.7) 34.9 (6.60–64.5) 4.5 (0.0–57) 2.95 (0.0–38.8) 0.8 (0.0–12.9)

Good mobilizers (n=10) Prior mobilization 0.25 (0.09–1.60) 0.38 (0.18–0.61) 6.95 (0.2–22.8) 24.88 (3.0–39.1) 26.95 (11.5–50.2) 4.94 (0.0–32.4) 1.6 (0–11.2) Study mobilization 1.45 (0.18–1.95) 1.06(0.49–2.18) 4.45 (0.8–13.1) 11.25 (0.5–25.9) 1.44 (0.31–5.00) 2.70 (0.0–12.3) 1.08 (0–18.6)

Numbers are median (range). were 4200  109/l (Po0.02) did so and the CD34+ yield 30 showed a significant correlation (Po0.001) to the baseline platelet counts. No significant association was seen for the filgrastim-alone group (r ¼À0.07).

- p= 0.15 20 CD34+ and subset measurements

+ The levels of CD34 cells in the good mobilizer group % CD38 (median 0.47%) were significantly higher in the study 10 mobilization compared with those of the prior mobilization (median 0.18%, P ¼ 0.002) and with the two mobilizations from the poor mobilizers (P ¼ 0.02 and 0.02, respectively). + 0 The levels of peripheral blood CD34 cells in the poor - rHuSCF + rHuSCF mobilizer group showed no difference between the prior and the study mobilization (median 0.14 vs 0.14%, P ¼ 0.72). + The CD34 cells from 18 patients who had sufficient 40 pilot tubes for coexpression analysis were studied (Table 2). Eight patients were from those whose yield from the 6 rHuSCF-stimulated mobilization did not reach the 2 Â 10 / 30 p<0.01 kg threshold (ie, poor mobilizers), and 10 were from those + whose yield from the rHuSCF-stimulated mobilization reached the threshold (ie, good mobilizers). 20 As defined by CD38, MDR, KDR, Thy-1, and c-kit, the % Thy-1 phenotypic subsets of CD34+ cells from the prior mobilization did not differ significantly between the poor 10 and good mobilizers. Hence, the phenotype of CD34+ cells mobilized by filgrastim did not predict the result of the subsequent mobilization with rHuSCF added. 0 In the good mobilizers, the median percentage of - rHuSCF + rHuSCF CD34+CD38- cells (4.5%) in the study mobilization was lower than in the prior mobilization (7.0%), but the 40 difference did not reach statistical significance (P ¼ 0.14). Fewer CD34+ cells coexpressed Thy-1 (median 11.3%) after mobilization with rHuSCF compared with the prior 30

+ p= 0.21 mobilization (24.9%; Po0.01), although the peak fluores- cence did not differ. There was no difference in the percentage of CD34+ cells expressing MDR and KDR 20

(Figure 3). % MDR When CD34+ cells from the study mobilization were 10 studied, the Thy-1 positivity was higher in the poor mobilizer group compared with the good mobilizer group (median 34.9 vs 11.3%, P ¼ 0.005). 0 - rHuSCF + rHuSCF Measurement of downregulation of c-kit (Figure 4) Figure 3 A comparison of the CD38, CD90, MDR expression on CD34+ + The phenotypes of the CD34 populations from the first cells between the prior and the study mobilizations in patients classified as and second mobilizations were compared. In the poor good mobilizers.

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 376

Good Mobilisers Poor Mobilisers

75 75

+ 50 + 50 p= 0.006 p=0.19 % c-kit % c-kit 25 25

0 0 - rHuSCF + rHuSCF - rHuSCF + rHuSCF

Good Mobilisers Poor Mobilisers

50 75

40 p=0.002 50 pk pk p=0.35 + 30 +

20 % c-kit % c-kit 25

10

0 0 - rHuSCF + rHuSCF - rHuSCF + rHuSCF

Figure 4 c-kit expression on CD34+ cells in poor and good mobilizers in the prior and the study mobilization. pk ¼ peak fluorescence. The line shows the median.

mobilizer group, no significant differences were observed; mobilization in nonrandomized studies, but this is the first that is, rHuSCF did not induce any change in the CD34+ report of a historically controlled study showing its efficacy in populations in this group (paired t-test). In contrast, in the patients who had failed prior mobilization. good mobilizer group, the level c-kit expression was Our findings show that the addition of rHuSCF is an significantly lower on CD34+ cells from the second alternative to filgrastim dose escalation in remobilization. mobilization based on both percent positive (median In a previous report we described that doubling the dose of 26.95 vs 1.44%, P ¼ 0.006) and peak fluorescence (22.4 filgrastim led to successful mobilization in six of 10 patients vs 10.8, P ¼ 0.002). Thus, successful mobilization with who had failed a prior mobilization with filgrastim- rHuSCF was associated with strong downregulation of plus-chemotherapy or filgrastim-only.13 Further, patients its receptor. enrolled in this filgrastim dose escalation program con- tinued to show a 50% success rate (data not shown). While a comparison between these two options is still Discussion to be performed, rHuSCF may enhance mobilization using nonfilgrastim-dependent pathways and may be This study shows that rHuSCF enhances peripheral blood useful even in patients who fail mobilization with stem cell mobilization in patients who had failed previously filgrastim. with either filgrastim-alone mobilization or chemotherapy- Stem cell factor is the ligand for the receptor tyrosine plus-filgrastim mobilization; 54 and 45%, respectively, of kinase, c-kit, which has important signalling functions in patients in these two groups reached the threshold dose. early hemopoietic cells leading to cell survival, proliferation When the yields of the prior and the study mobilizations were and differentiation.18 Downregulation of c-kit has been one pooled, 77 and 71%, respectively, of patients in these two of the most consistent findings on mobilized CD34+ groups reached the 2 Â 106 CD34+ cells/kg threshold and cells,14,15,19 particularly on cells mobilized with rHuSCF,14 were enabled to proceed to transplant. Furthermore, for a and was shown to correlate with the yield of mobiliza- minimum prior mobilization yield of 0.5 Â 106 CD34+ cells/ tion.4,15 This study supports that such a downregulation is kg, no clear threshold existed below which a second rHuSCF- part of successful mobilization with rHuSCF in patients supported mobilization would not be successful. This result is with failed prior mobilization. consistent with data that Azar et al17 reported in abstract Binding of SCF downregulates cell surface c-kit by form, showing a 34% success rate in remobilization when promoting dimerization, ubiquitination, internalization rHuSCF was administered to 67 patients who failed a prior and degradation within minutes of binding.20 In contrast, mobilization. rHuSCF has been shown to enhance filgrastim mobilization by rHuSCF and other agents takes several

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 377 days, suggesting that additional mechanisms may be References operative. c-kit mRNA is known to be downregulated by other cytokines, for example, granulocyte–macrophage 1 To LB, Sheppard KM, Haylock DN et al. Single high dose of colony-stimulating factor (GM-CSF) and transforming cyclophosphamide enable the collection of high numbers of 21,22 haemopoietic stem cells from the peripheral blood. Exp growth factor (TGF)b. Hence, modulation of cell Hematol 1990; 18: 442–447. surface c-kit may reflect the sequence of complex changes 2 Sheridan WP, Begley CG, Juttner CA et al. Effect of peripheral during the progress of mobilization. blood progenitor cells mobilized by filagrastim (G-CSF) on The consistent association of reduced c-kit expression in platelet recovery after high dose chemotherapy. Lancet 1992; CD34+ cells mobilized in different ways suggests that it 339: 640–644. may be central to the mechanism of mobilization. The 3 Hoglund M, Smedmyr B, Simonsson B et al. Dose dependant importance of the SCF/c-kit ligand pair in mobilization is mobilization of haematopoietic progenitor cells in healthy further indicated by a number of studies in animals. Cynshi volunteers receiving glycosylated rHuG-CSF. Bone Marrow et al23 showed that Sl/Sld mice that lack membrane- Transplant 1996; 18: 19–27. associated SCF, or W/Wv mice that express only kinase- 4 To LB, Haylock DN, Simmons PJ, Juttner CA. The biology and clinical uses of blood stem cells. Blood 1997; 89: 2233– c-kit deficient on their stem cells, mobilize less well with 2258. granulocyte colony-stimulating factor (G-CSF). These 5 Broudy VC. Stem cell factor and hematopoiesis. Blood 1997; studies suggest that triggering of c-kit signalling by 90: 1345–1364. membrane SCF on stromal cells is required for optimum 6Moskowitz CH, Stiff P, Gordon MS et al. Recombinant G-CSF-induced mobilization. methionyl human stem factor and filgrastim for peripheral Binding of c-kit on hemopoietic cells to membrane SCF blood progenitor cell mobilization and transplantation in non- on stromal cells can directly mediate adhesion.24,25 It is Hodgkins lymphoma patients: results of a phase I/II trial. possible that rHuSCF could contribute to mobilization by Blood 1997; 89: 3136–3147. competing with membrane SCF in this interaction. How- 7 Begley CG, Basser R, Mansfield R et al. Enhanced levels and ever, the time course of mobilization and the failure of clonogenic capacity of blood progenitor cells following administration of SCF plus G-CSF to humans. Blood 1997; various mobilization regimens in W/Wv mice appear 90: 3378–3389. incompatible with this simple explanation. Furthermore, 8 Shpall EJ, Wheeler CA, Stewart A et al. A randomised phase 3 it has been suggested that cytokine-mediated progenitor cell study of peripheral blood progenitor cell mobilization with 14,26 proliferation and mobilization may be linked, in which stem cell factor and filgrastim in high-risk breast cancer case the known synergism between rHuSCF and filgrastim patients. Blood 1999; 93: 2491–2501. in proliferation could also enhance mobilization. It is also 9 Weaver A, Chang J, Wrigley E et al. Randomised comparison possible that SCF may interact with integrins another of progenitor-cell mobilization using chemotherapy, stem-cell family of molecules involved in mobilization27 but this is factor, and filgrastim or chemotherapy plus filgrastim alone in beyond the scope of the study. patients with ovarian cancer. J Clin Oncol 1998; 16: 2601–2612. In this study, the expression of marker of early ontogeny 10 Basser RL, To LB, Begley G et al. Rapid haemopoeitic on CD 34+ cells did not show any distinctive difference recovery after multiple high dose chemotherapy: enhancement of filgrastim-induced progenitor cell mobilization by recombi- between the failed and the successful mobilizations, except nant human stem cell factor. J Clin Oncol 1998; 16: 1899–1908. for a reduction in Thy-1 expression. The significance of 11 Stiff P, Gingrich R, Luger S et al. A randomised phase 2 study such an isolated finding is uncertain. There was certainly no of PBPC mobilization by stem cell factor and filgrastim in evidence of delayed or failed engraftment in patients heavily pretreated patients with Hodgkin’s disease or non- transplanted with rHuSCF-mobilized cells. Hodgkins lymphoma. Bone Marrow Transplant 2000; 26: 471– In conclusion, rHuSCF seems capable of enhancing 481. filgrastim-alone or chemotherapy-plus-filgrastim mobiliza- 12 Lie AKW, Rawling TP, Bayly JL et al. Progenitor cell yield in tions even in patients who had failed prior mobilization. sequential blood stem cell mobilization in the same patients: The mobilized cells are capable of rapid hemaopoietic re- insights into chemotherapy dose escalation and combination of constitution. Downregulation of c-kit on CD34+ cells is haemopoietic growth factor and chemotherapy. Br J Haematol 1996; 95: 39–44. again noted in successful mobilization with the addition of 13 Lie AKW, Hui C-H, Rawling T et al. Granulocyte colony- rHuSCF in these patients with a failed prior mobilization. stimulating factor (G-CSF) dose-dependent efficacy in periph- rHuSCF and c-kit downregulation may enhance mobili- eral blood stem cell mobilization in patients who had failed zation via proliferation, integrin and other signalling initial mobilization with chemotherapy and G-CSF. Bone pathways. Marrow Transplant 1998; 22: 853–857. 14 Roberts MM, Swart BW, Simmons PJ et al. Prolonged release and c-kit expression of haemopoietic precursor cells mobilized by stem cell factor and granulocyte colony stimulating factor. Acknowledgements Br J Haematol 1999; 104: 778–784. 15 Simmons PJ, Leavesley DI, Levesque J-P et al. The mobiliza- We thank Diane Landorf for secretarial assistance, Melanie tion of primitive haemopoietic progenitors into the peripheral Seeger for literature searches, and Mary-Ann Foote for blood. Stem Cells 1994; 12(Suppl. 1): 187–202. reviewing the manuscript. The following research nurses 16Ziegler BL, Valteri M, Porada GA et al. KDR receptor: a key contributed to trial coordination and data collection: Marjorie marker defining haematopoietic stem cells. Science 1999; 285: Goward, Kerry Kristaly, Kathy Lape, Rosetta Maisano, Jenny 1553–1558. Muirhead, Caroline Rawling, Leonie Snowden, and Sonya 17 Azar N, Vantelon JM, Hashmanti P et al. Ancestim(r- Stephens. metHuSCF): mobilization of peripheral blood stem cells

Bone Marrow Transplantation Stem cell factor for failed mobilization LB To et al 378 (PBSC) in patients with prior unsuccessful mobilization with stability and cell-surface protein expression in hematopoietic G-CSF. Experience of 67 cases of compassionate use in progenitors. Blood 1994; 83: 3138–3145. France. Blood 2000; 96: 176a. 23 Cynshi O, Satoh K, Shimonaka Y et al. Reduced response to 18 Ashman LK. The biology of stem cell factor and its receptor granulocyte colony-stimulating factor in W/W and SL/SL C-kit. Int J Biochem Cell Biol 1999; 31: 1037–1051. mice. Leukaemia 1991; 5: 75–80. 19 To LB, Haylock DN, Dowse T et al. A comparative study of 24 Kaneko Y, Takenawa J, Yoshida O et al. Adhesion of mouse the phenotype and proliferative capacity of peripheral blood mast cells to fibroblasts: adverse effects of steel (S1) mutation. (PB) CD34+ cells mobilized by four different protocols and J Cell Physiol 1991; 147: 224–230. those of steady-phase PB and bone marrow CD34+ cells. 25 Avraham H, Scadden DT, Chi S et al. Interaction of human Blood 1994; 84: 2930–2939. bone marrow fibroblasts with megakaryocytes: role of the c-kit 20 Yee Ns, Hsiau CW, Serve H et al. Mechanism of down- ligand. Blood 1992; 80: 1679–1684. regulation of c-kit receptor. Roles of receptor tyrosine kinase, 26Morrison SJ, Wright DE, Weissman IL. Cyclophosphamide/ phosphatidylinositol 30-kinase, and protein kinase C. J Biol Chem 1994; 269: 31991–31998. granulocyte colony-stimulating factor induces haematopoietic 21 Caruana G, Ashman LK, Fujita J, Gonda TJ. Responses of stem cells to proliferate prior to mobilization. Proc Natl Acad the murine myeloid cell line FDC-P1 to soluble and Sci USA 1997; 94: 1908–1193. membrane-bound forms of steel factor (SLF). Exp Hematol 27 Papayannopoulou T, Priestley GV, Nakamoto B. Anti VLA4/ 1993; 21: 761–768. VCAM-1-induced mobilization requires cooperative signalling 22 Dubois CM, Ruscetti FW, Stankova J, Keller JR. Transform- through the kit/mkit ligand pathway. Blood 1998; 91: 2231– ing growth factor-beta 2 microglobulin regulates c-kit message 2239.

Bone Marrow Transplantation