peripheral blood stem cell kawano

ORIGINAL ARTICLE

Peripheral blood stem cell mobilization by granulocyte colony-stimulatingfactor alone and engraftment kinetics following autologous transplantationin children and adolescents with solid tumor

H Watanabe1, T Watanabe1, H Suzuya1, Y Wakata1, M Kaneko1, T Onishi1, Y Okamoto2, T Abe1,Y Kawano2, S Kagami1 and Y Takaue3

1Department of Pediatrics, University of Tokushima Graduate School of Medical Science, Tokushima, Japan; 2Department ofPediatrics, University of Kagoshima Graduate School of Medical Science, Kagoshima, Japan and 3Department of Stem Cell Therapy,National Cancer Center, Tokyo, Japan

In 56 pediatric and adolescent patients (median age 7years, range 121) with various solid tumors, peripheralblood stem cells (PBSC) were mobilized with granulocytecolony-stimulating factor (G-CSF) alone, and the yields ofPBSC and engraftment kinetics following autologousperipheral blood stem cell transplantation (PBSCT) wereevaluated retrospectively. Granulocyte colony-stimulatingfactor (10 lg/kg) was injected subcutaneously for mobi-lization when patients showed no inuence of previouschemotherapy, and administration was continued for 5days. The peaks of CD34 cells and colony-formingunits-granulocyte/macrophage in the blood were observedon days 4 through 6 of G-CSF administration in allpatients. Peripheral blood stem cell harvest was com-menced on day 5 of G-CSF treatment. Compared to theresults in patients mobilized by chemotherapy plus G-CSF(N 18), the progenitor cell yields were lower in patientsmobilized with G-CSF alone. However, there were nosignicant differences in WBC and ANC engraftmentcompared to the chemotherapy plus G-CSF mobilizationgroup. Platelet recovery following autologous PBSCTwas delayed in patients mobilized with G-CSF alone. Themedian time taken for ANC and platelet counts to reach0.5 109 and 20 109/l was 12 days (range: 928) and 15days (855), respectively, in all patients who receivedPBSC mobilized by G-CSF alone. In summary, mobiliza-tion with G-CSF alone can mobilize sufcient CD34

cells for successful autografting and sustained hematolo-gical reconstitution in pediatric and adolescent patientswith solid tumors, and even in heavily pre-treated patients.Bone Marrow Transplantation (2006) 37, 661668.doi:10.1038/sj.bmt.1705304; published online 20 February2006Keywords: PBSC; mobilization; G-CSF; engraftment;pediatric and adolescent solid tumor

Introduction

Autologous hematopoietic stem cell transplantation isincreasingly performed for the treatment of variouspediatric and adolescent cancers,1,2 and currently mostcenters use peripheral blood stem cells (PBSC) as thesource of stem cells to support high-dose chemotherapy.The advantages of PBSC include rapid hematopoieticrecovery following transplantation, and less of a potentialfor tumor cell contamination of the grafts.3 The doseof PBSC administered determines the success of engraft-ment,4 which makes effective PBSC mobilization a criticalstep for the entire procedure. Methods for PBSC mobili-zation that are available for cancer patients includechemotherapy-induced mobilization, and mobilization byhematopoietic growth factor with or without chemo-therapy.58

Although chemotherapy plus granulocyte colony-stimu-lating factor (G-CSF) is a standard mobilization regimenfor patients with cancer,9 there is great variability inpredicting the optimal day to start collecting PBSC. Inparticular, it takes longer to reach the peak of circulatingCD34 cells following pre-mobilization chemotherapy,especially in heavily pre-treated patients. Although thedaily measurement of blood CD34 cells from the day ofrising WBC to the day of apheresis has been used todetermine the time to start collection,10 the frequentmeasurement of CD34 cells is both time consuming andlabor intensive. We sometimes miss the opportunity forcollection when WBC is rising very slowly. Furthermore,owing to a paucity of technical staff, we seek to avoidcollection procedures on weekends. These considerationsmotivated us to adopt mobilization with G-CSF alone forpatients undergoing autologous transplantation. Mobiliza-tion with G-CSF alone is exclusively used for healthydonors, and the mobilization kinetics have been wellstudied.11 Whether or not mobilization with chemotherapyplus G-CSF is preferable to mobilization with G-CSF aloneremains to be established in the autologous setting.Mobilization with G-CSF alone in cancer patients mighthave advantages, such as ease of making a collectionschedule without a daily determination of CD34 cells in

Received 11 November 2005; revised 5 December 2005; accepted 4January 2006; published online 20 February 2006

Correspondence: Dr T Watanabe, Department of Pediatrics, Universityof Tokushima Graduate School of Medical Science, Kuramoto-cho 2-18-15, Tokushima 770-8503, Japan.E-mail: [email protected]

Bone Marrow Transplantation (2006) 37, 661668& 2006 Nature Publishing Group All rights reserved 0268-3369/06 $30.00

www.nature.com/bmt

the blood, and the avoidance of neutropenic fever andadditional transfusion. In addition, mobilization withG-CSF alone can be performed on an outpatient basis.In this study, we examined the effectiveness of PBSC

mobilization with G-CSF alone in pediatric and adolescentpatients with solid tumors, including heavily treated patients.

Materials and methods

PatientsData from 74 patients who were 21 years of age or youngerand who underwent autologous PBSC harvest for primarycancer treatment between April 1997 and June 2005 wereanalyzed retrospectively. Their characteristics are shown inTable 1. The patients, who were mobilized with G-CSFalone (N 56), were categorized according to prior treat-ment: patients who received four or fewer cycles ofchemotherapy with/without local irradiation (Group 1;N 21); patients who received more than four cycles ofchemotherapy, or three or more cycles of chemotherapywith extended irradiation such as cerebrospinal irradiation(Group 2; N 23); and patients who have been mobilizedpreviously and received high-dose chemotherapy with PBSCsupport following conventional chemotherapy, and thenthey had a recurrent disease, and mobilization wasattempted again (Group 3; N 12). In Group 3, the mediantime from the last high-dose chemotherapy to PBSC harvestwas 8.5 months (range: 348). Patients who were mobilizedwith chemotherapy followed by G-CSF (N 18) during thesame period in our institute were also evaluated as a controlgroup. Groups 13, and the control group were comparablein age. The clinical protocols for peripheral blood stem celltransplantation (PBSCT) were approved by the InstitutionalReview Board at Tokushima University Hospital. Writteninformed consent was obtained from the patients (if thepatient was more than 10 years old) or their guardians.

Peripheral blood stem cell mobilization and harvestFor mobilization with G-CSF alone, G-CSF was initiatedduring the steady state, when blood cells had recoveredfrom the inuence of chemotherapy. The time from the lastday of chemotherapy to the commencement of G-CSFaveraged 19 days (range: 035; 0 means that patients didnot receive chemotherapy before mobilization) in Group 1,19 days (1141) in Group 2 and 29 days (1585) in Group 3(Table 1). Granulocyte colony-stimulating factor (10 mg/kg)was administered for 5 consecutive days. Principally,G-CSF was injected subcutaneously once a day at 0900each day. Blood samples for daily analysis of completeblood count, colony-forming unit-granulocyte/macrophage(CFU-GM) assay and CD34 cell assay were drawnimmediately before the injection of G-CSF. On day 5 ofG-CSF treatment, after the fth injection of G-CSF, PBSCharvest was commenced using a continuous cell separator.For mobilization with chemotherapy plus G-CSF, the

patients received disease-oriented chemotherapy; G-CSFwas started when WBC reached a nadir, and completeblood counts were followed. The dose of G-CSF was5 mg/kg intravenously, and was continued daily until the

completion of PBSC harvest. We usually started to harvestPBSC when the platelet count reached over 100 109/lwithout transfusion support, with a concomitant risingWBC. Chemotherapeutic drugs and dosages used formobilization were etoposide 500mg/m2 and cisplatin90mg/m2 (N 8), carboplatinum 800mg/m2 and etoposide375mg/m2 (N 5), doxorubicin 75mg/m2 and ifosphamide7.5 g/m2 (N 3), and high-dose cytosine arabinoside20 g/m2 (N 2). Circulating CD34 cells and CFU-GMwere not determined in this group.Apheresis was usually initiated at 1000 and 150300ml/kg

(max. 10 l) was processed per session. The details of theharvesting procedure have been described previously.12

Briey, mostly in children less than 10 years old, blood wasdrawn from a radial artery using a 2024 gauge catheterand given back to an antecubital vein using an 1824 gaugecatheter. For donors weighing less than 20 kg, the extra-corporeal circuit was pre-primed with donated WBC-depleted blood, depending on the machine and separatingchambers. In adolescents, we used antecubital veins tocollect PBSC. Calcium gluconate was continuously infusedat a dose of 1ml per processed 200ml blood duringapheresis. The target doses were set at42 106/kg CD34cells for one course of high-dose therapy. In 19 of 56patients who were mobilized with G-CSF alone, doublePBSCT was planned, along with repeated aphereses tosupport two rounds of high-dose chemotherapy. Twoaphereses were performed on day 5 in most patients, andadditional aphereses were performed on days 6 and 7.Peripheral blood stem cells were cryopreserved in 10%dimethylsulfoxide in a deep freezer (1501C) until use.Mobilization failure was dened as a patient who never

reached a CD34 cell level in the blood of 20 106/l, or apatient from whom a proper CD34 cell level was notcollected in the rst apheresis product, and in such cases wedid not attempt further apheresis.

Flow cytometer analysis for CD34 cellsOne hundred microliters of cell suspension were added to atest tube containing isotype control (phycoerythrin-mouseIgG1) and phycoerythrin-conjugated CD34 monoclonalantibody (anti-HPCA2 antibody; Becton Dickinson, SanJose, CA, USA) at a concentration of 1mg antibody/106

cells. MNCs from samples were incubated for 30min, andthen washed twice. Samples were analyzed with a FACScanow cytometer (Becton Dickinson). After the function wasveried, samples were drawn into the ow cytometer usingforward scatter and side scatter, as gating parameters,along with debris subtraction techniques to determine thecharacteristics of the cells. A total of 20 000 events werecounted to identify the mononuclear cell fraction. The owcytometric data were analyzed using a gated analysis via aset of SSC-FL parameters for CD34 cells to calculate thepercentage of positive cells.The number of circulating CD34 cells was calculated

using the following formula:

CD34 cells=l WBC 106=lpercentage of MNC %=100frequency of positive cells %=100

PBSC mobilization by G-CSF alone in pediatric cancer patientsH Watanabe et al

662

Bone Marrow Transplantation

Table 1 Patients characteristics

Group 1 (N 21) Group 2 (N 23) Group 3 (N 12) Total G-CSF alone (N 56) Chemotherapy plus G-CSF (N 18)

Age: (Median (range)) 7 (121) 8 (121) 6 (213) 7 (121) 7 (121)Sex (M/F) 14/7 18/5 10/2 42/14 13/5Primary tumor Germ cell tumor 6 Medulloblastoma 8 Medulloblastoma 5 Medulloblastoma 16 Medulloblastoma 7

Medulloblastoma 3 Neuroblastoma 4 Neuroblastoma 3 Germ cell tumor 13 Neuroblastoma 3Pilocytic astrocytoma 3 Germ cell tumor 4 Germ cell tumor 2 Neuroblastoma 5 NHL 3

PNET/Ewings 3 Wilms Tumor 2 Pilocytic astrocytoma 1 Pilocytic astrocytoma 4 Rhabdomyosarcoma 3Rhabdomyosarcoma 2 Osteosarcoma 2 Wilms tumor 1 Wilms tumor 3 PNET/Ewings 1

Pontine glioma 1 NHL 1 Osteosarcoma 3 Osteosarcoma 1Adrenal cancer 1 MRK 1 PNET/Ewings 3Retinoblastoma 1 Hepatoblastoma 1 Rhabdomyosarcoma 2Ependymoma 1 Others 10

Metastasis to bone marrow 1 1 2 4 2Prior chemotherapyPlatinum-based 14 11 0 25 4Alkylator-Based 5 3 0 8 1Platinum/Alkylator-based 2 9 12 23 6No. of cycles 3 (23) 10 (412) 8 (315) 6 (215) 5 (310)No. of high-dose therapies 0 0 2 (13) 2 (13) 0

Local irradiation 4 1 0 5 1Extensive irradiation 0 5 5 10 5Time from onset to PBSC harvest (months) 4 (113) 7 (425) 19 (948) 6 (130) 5 (315)Last chemotherapy to G-CSF (days) 19 (035) 19 (1141) 29 (1585) 24 (085) 18 (1126)

Types of G-CSFFilgrastim 9 11 6 26 10Lenograstim 10 10 6 26 8Nartograstim 2 2 0 4 0

Abbreviations: G-CSF granulocyte colony-stimulating factor; MRKmalignant rhabdoid tumor; NHL non-Hodgkins Lymphoma; PBSC peripheral blood stem cell; PNET primitive neuroectodermaltumor.

PB

SCm

obilizationby

G-CSF

alonein

pediatriccancer

patientsH

Watanabe

etal

663

Bo

ne

Marro

wT

ran

spla

nta

tion

Colony-forming units-granulocyte/macrophage assayColony-forming units-granulocyte/macrophage were deter-mined using MethoCult GF H4434V (VERITAS Co.,Tokyo, Japan) according to the manufacturers guidelines.Briey, MNCs were incubated in methylcellulose mediumsupplemented with recombinant human erythropoietin,recombinant human stem cell factor, recombinant humanGM-CSF, recombinant human G-CSF and recombinanthuman IL-3. Cells were placed in 35-mm culture dishes intriplicate and incubated in an ESPEC N2O2CO2, a,m,o-110 incubator (Tabai ESPEC Co., Osaka, Japan), whichmaintained a humid atmosphere of 5% carbon dioxide, 5%oxygen and 90% nitrogen. After 14 days of incubation, thenumber of CFU-GM was counted under an invertedmicroscope. The number of circulating CFU-GM wascalculated with the following formula:

CFUGM=l WBC 106=lMNC %=100no: of colonies per well

no: of scattered MNC per well

High-dose chemotherapy and autologous peripheral bloodstem cell transplantationThe high-dose regimen depended on the type of cancer.Mostly, regimens consisted of 200mg/kg cyclophosph-amide and 1600mg/m2 carboplatinum for pilocytic astro-cytoma and medulloblastoma, 750mg/m2 ThioTEPA,450mg/m2 MCNU and 1500mg/m2 cyclophosphamidefor medulloblastoma, 1600mg/m2 carboplatinum and1500mg/m2 etoposide for germ cell tumor, 1600mg/m2

etoposide, 750mg/m2 ThioTEPA and 180mg/m2 melpha-lan for rhabdomyosarcoma and peripheral neuroecto-dermal tumor, 750mg/m2 ThioTEPA and 180mg/m2

melphalan for osteosarcoma, 1600mg/m2 carboplatinum,1500mg/m2 etoposide and 180mg/m2 melphalan forneuroblastoma and 450mg/m2 MCNU, 16 g/m2 cytosinearabinoside (Ara-C), 1600mg/m2 etoposide and 100mg/kgcyclophosphamide for non-Hodgkins lymphoma. Posttransplant G-CSF use was limited to patients who showedpoor PBSC collection, such as Group 3 patients or forpatients, who were complicated with severe infection.

Statistical analysisStatistical analysis was performed using SPSS statisticalsoftware (SPSS Co., Tokyo, Japan). Groups were com-pared using the non-paired t-test. The relationshipsbetween different hematological parameters and PBSCyields were analyzed by the Spearman rank correlation test.Pearsons sample correlation coefcient and the corre-sponding P-value were calculated for the null hypothesis ofno correlation.

Results

Circulating cell kinetics during mobilization withgranulocyte colony-stimulating factor alone (Figure 1)The changes in WBC, platelet, circulating CD34 cells andCFU-GM in patients mobilized with G-CSF alone are

shown in Figure 1. The median WBC before the initiationof G-CSF was 3.35 109/l (range: 1.207.90). The medianhemoglobin level and platelet count before the initiation ofG-CSF were 9.6 g/dl (7.413.5) and 249 109/l (70647),respectively. WBC increased rapidly after G-CSF treat-ment, with a peak of 29.67 109/l (range: 9.3062.00) onday 6. The platelet count decreased gradually duringG-CSF treatment, but this decrease was not signicant.After apheresis on day 5, the platelet count signicantlydecreased to 144 109/l (44410). The kinetics of circula-

B

B

BB

B

B

0

5

10

15

20

25

30

35

WB

C

BB

B

B

B

B

0

10

20

30

40

50

60

70

80

CD

34+

cells

B

B

B

B

B

B

0

2

4

6

8

10

12

CFU

-GM

Day 6Day 5Day 4Day 3Day 2Day 1



Days of G-CSF administration

a

b

c

Figure 1 Circulating cell kinetics during granulocyte colony-stimulatingfactor mobilization. (a) WBC ( 109/l), (b) circulating CD34 cells( 106/l), (c) circulating colony-forming units-granulocyte/macrophage( 106/l).


664


tory CD34 cells and CFU-GM during mobilization weresynchronized, and both of the peak values occurred on days4 through 6 of G-CSF treatment. The numbers of CD34

cells and CFU-GM gradually increased between days 1 and3, and markedly increased on days 4 and 6 following theinitiation of G-CSF treatment. The peak CD34 cell levelwas observed on day 5 in 37 of 56 patients (66%), on day 4in 10 patients (18%) and on day 6 in nine patients (16%).The peak CFU-GM level was observed on day 5 in 40 of 56patients (71%), on day 4 in seven patients (13%) and onday 6 in nine patients (16%). The median numbers ofcirculating CD34 cells and CFU-GM on day 5 were62.4 106/l (4.13192.60) and 11.16 106/l (0.05648.98),respectively.

Mobilization effects and peripheral blood stem cellapheresis yieldsThe results of mobilization and apheresis are shown inTable 2. Peripheral blood stem cell yields were calculated asnumber per unit of processed blood (l) per unit weight ofthe patient (kg). The peak values of circulating CD34cells and CFU-GM in Group 3 were signicantly lowerthan those in Groups 1 (P 0.002) and 2 (P 0.049).Compared to patients mobilized by chemotherapy plusG-CSF, the CD34 cell yields were signicantly lower inall of the patients mobilized with G-CSF alone (P 0.045).Compared to patients mobilized by chemotherapy plusG-CSF, the CD34 cell yields were signicantly lowerin Group 3 patients (P 0.014). Compared to patients

mobilized by chemotherapy plus G-CSF, the CFU-GMyields were signicantly lower in Group 2 (P 0.033). TheCD34 cell yield was signicantly lower in Group 3compared to Group 1 (P 0.014). There were no differ-ences in CD34 cell yields between Groups 1 and 2, andthe control group. The CFU-GM yield in Group 1 wassignicantly higher than that in Groups 2 and 3 (P 0.033and 0.048, respectively).We analyzed several factors that might be used to

predict PBSC yields such as age and hematologicalparameters before and during G-CSF treatment in all ofthe patients who were mobilized with G-CSF alone. In allof these patients, the number of circulating CD34 cells onday 5 was signicantly correlated with the CD34 yield(Figure 2). There was no signicant correlation betweenany other factors and PBSC yields.

Toxicity of the mobilization procedureNo patient required a reduced dose of G-CSF. The adverseeffects induced by G-CSF injection were tolerable in allgroups. Although patients mobilized with G-CSF alonewho were older than 10 years sometimes complainedof slight fever, bone pain, lumbago or mild headache, thiswas sustained with or without medication. Clinical signsof severe infection were not evident in patients mobilizedwith G-CSF alone. However, among patients mobilizedwith chemotherapy plus G-CSF, almost all of the patientsexperienced neutropenia, and the nadir ANC was 0.10.4 109/l. There were episodes of fever during neutropenia

Table 2 Mobilization and collection of PBSC by apheresis

Group 1 (N 21) Group 2 (N 23) Group 3 (N 12) All (N 56) Chemotherapy plus G-CSF (N 18)

CD34+ cell( 106/)l (peak) NEMean 72.6 81.3 19.1 62.4Median (range) 82.9 (34.3118.5) 46.3 (14.0361.2) 16.5 (4.240.0) 34.3 (4.2361.2)

CFU-GM ( 106/)l (peak) NEMean 16.6 10.6 2.8 10.3Median (range) 9.3 (0.048.9) 7.8 (0.132.6) 0.9 (0.010.4) 3.3 (0.048.9)

Apheresis number*Mean 2.8 3.0 3.1 2.9 2.7Median (range) 3 (24) 3 (24) 3 (24) 3 (24) 2 (15)

MNC/kg/l collectedMean 0.94 1.08 1.12 1.02 0.69Median (range) 0.61 (0.182.10) 0.67 (0.126.26) 1.05 (0.282.59) 0.67(0.126.26) 0.64 (0.091.39)

CD34+ cells/kg/l collectedMean 1.33 1.19 0.42* 1.06* 2.62Median (range) 0.94 (0.073.82) 0.83 (0.087.78) 0.36 (0.081.39) 0.60 (0.077.78) 0.67 (0.0113.88)

CFU-GM/kg/l collectedMean 2.01 1.09* 1.35 1.37* 3.71Median (range) 1.54 (0.215.01) 0.77 (0.103.56) 0.68 (0.054.15) 0.82 (0.055.01) 0.70 (0.0620.25)

Mobilization failure (no) 1 2 1 4 1

Abbreviations: CFU-GM colony-forming units-granulocyte/macrophage; G-CSF granulocyte colony-stimulating factor; PBSC peripheral blood stemcells.PBSC were collected to support double sequential high-dose chemotherapy in three patients in Group 1, 12 patients in Group 2, four patients in Group 3 andno patients in the control group.Number was divided by the target number of planned high-dose chemotherapies.*Signicantly low compared to the control group.


665


in 13 of 18 patients (72%) mobilized with chemotherapyplus G-CSF, which required treatment with intravenousantibiotics and/or antifungal agents. Culture was positivefor ve patients who were mobilized with chemotherapyplus G-CSF.

EngraftmentThe results of engraftment following autologous PBSCTare shown in Table 3. In Group 1, high-dose chemotherapywas cancelled because of progressive disease in one. InGroup 2, high-dose chemotherapy was cancelled because

of stable disease in one and poor PBSC collection in one.In Group 3, only one patient declined high-dose chemo-therapy because of poor PBSC collection, and proceeded toallogeneic cord blood cell transplantation.Three patients in Group 1, 12 in Group 2 and four in

Group 3 received double sequential high-dose chemo-therapy. None of the patients in the control group receiveddouble transplantation. There were no signicant differ-ences in WBC and ANC engraftment among all groups,although platelet recovery was delayed in patients mobi-lized with G-CSF alone, especially Group 3 patients. Graftfailure was not observed in any of the transplanted patients.

Discussion

This study was a retrospective analysis of pediatric andadolescent patients with solid tumors treated with auto-logous PBSCT at our institution. Therefore, the disease andconditioning regimens are heterogeneous. In addition, therelatively small number of patients in some treatmentgroups and the retrospective analysis may make it difcultto draw any denitive conclusions. Nevertheless, our datasuggest that G-CSF alone may be capable of mobilizinga sufcient number of PBSC for successful autograftingafter high-dose chemotherapy in pediatric and adolescentpatients with solid tumors. The study of mobilizationkinetics showed a similar pattern of PBSC mobilizationwith allogeneic donors, and the peaks of circulatingCD34 cells and CFU-GM were observed on days 4through 6 of G-CSF treatment. Thus, when PBSC aremobilized by G-CSF alone, we can easily determine the

1

0

1

2

3

4

5

6

7

8

CD

34+

cell

yiel

d

50 0 50 100 150 200 250 300 350 400Circulating CD34+ cells on day 5

Figure 2 Correlation between circulating CD34 cells ( 106/l) on day 5of granulocyte colony-stimulating factor (G-CSF) administration andCD34 cell yields in patients mobilized with G-CSF alone (N 38).CD34 cell yields were calculated as number per unit of processed blood (l)per unit weight of the patient (kg).

Table 3 Engraftment results

Group 1 (N 21) Group 2 (N 23) Group 3 (N 12) All (N 56) Chemotherapy plusG-CSF (N 18)

Total no. of transplants 24 35 16 75 18No. of cases of double transplant 3 12 4 19 0

No. of transplanted cellsMNC ( 108/kg)Mean 4.79 3.33 4.35 4.01 5.36Median (range) 4.03 (1.3013.20) 2.59 (1.4013.70) 5.58 (0.916.19) 3.20 (0.9113.70) 4.60 (2.1313.2)

CD34+ cells ( 106/kg)Mean 10.36 4.10* 2.43* 6.39 13.91Median (range) 4.51 (1.2048.20) 3.38 (0.369.65) 1.47 (0.415.90) 3.8 (0.3648.2) 8.7 (4.4674.5)

CFU-GM ( 105/kg)Mean 11.02 4.04* 0.99* 5.54* 17.77Median (range) 3.60 (0.3014.00) 3.65 (0.1012.80) 0.42 (03.74) 3.49 (014.00) 14.85 (0.5590)

Engraftment days (median (range))WBC41.0 109/l 11 (826) 12 (721) 12 (928) 11 (828) 12 (916)ANC40.5 109/l 11 (826) 13 (922) 12 (928) 12 (928) 12 (916)Platelet420 109/l 12 (728) 15 (831) 22** (855) 15 (855) 10 (716)Platelet450 109/l 16 (1056) 27** (849) 27* (20149) 20** (8149) 14 (728)

Post transplant G-CSF use (no.) 6 16 14 36 3

Abbreviations: CFU-GM colony-forming units-granulocyte/macrophage; G-CSF granulocyte colony-stimulating factor.*Signicantly low compared to the control group.**Signicantly delayed compared to the control group.


666


timing of commencing PBSC harvest without monitoringthe daily CD34 cell concentration. In contrast, inmobilization with chemotherapy plus G-CSF, it might bedifcult to determine the optimal timing of PBSC harvestfollowing chemotherapy because of the wide range ofoptimal mobilization time points.9 In mobilization withchemotherapy plus G-CSF, suggested predictive factors forinitiating harvests have included the peripheral WBC count,circulating CD34 cells, the peripheral absolute monocytecount and the platelet count.10,13 The measurement ofCD34 cells in peripheral blood before PBSC harvest hasbegun to be used, but it is too labor intensive to determinethis value every day around the time of PBSC harvest.Furthermore, when chemotherapy is used for PBSCmobilization, myelosuppression induced by chemotherapyis inevitable, and we might repeat unwanted chemotherapysimply to mobilize PBSC, and thus cause additional costsfor patient care. Most patients who received chemotherapyfor PBSC mobilization required treatment of neutropenicfever,14 as in our study. Drawbacks for mobilization withG-CSF alone include a 1-week delay of scheduled chemo-therapy and a lack of antitumor effects. However, if westarted to harvest after two or three courses of chemo-therapy instead of in the early phase of chemotherapy, thisdelay might not affect the outcome of the disease. Althoughthe dose of G-CSF is higher in mobilization with G-CSFalone than in post-chemotherapy mobilization, the durationof G-CSF administration can be much longer in post-chemotherapy administration than in mobilization withG-CSF alone. The total dose of G-CSF in post-chemo-therapy mobilization might be greater than in mobilizationwith G-CSF alone. Thus, mobilization with G-CSF alonemight offer clinical and economic advantages, and mightincrease convenience and improve the QOL of patients.It is often difcult to obtain the minimum requirement of

CD34 cells for safe engraftment after autologous PBSCTin heavily pre-treated patients. In adult lymphoma patients,predictors of poor mobilization were more than two priortreatment regimens and the WBC count on the rst dayof apheresis.15,16 In another study, patients who failedto mobilize had received more than seven cycles, and theauthors advocated that PBSC collection should be plannedearly in the course of chemotherapy.17 In adult cancerpatients, factors that predict a higher likelihood of poormobilization include increasing numbers of cycles of priorchemotherapy, prior radiation therapy and the presenceof overt marrow metastasis.1820 Autologous donors areassumed to have sustained an injury to the hematopoieticstem cell system that is responsible for the poor mobiliza-tion. Limited data are available with regard to PBSCcollection with G-CSF alone in pediatric patients withcancer.21 We showed that PBSC could be mobilized byG-CSF alone even in most of the heavily treated pediatricpatients with solid tumors using the same dose as in healthydonors. Most of the patients received platinum-based oralkylator-based chemotherapy before PBSC harvest. Wecould not analyze the effects of specic chemotherapeuticagents on PBSC mobilization. We merely analyzed theeffects of the degree of previous chemotherapy on PBSCmobilization. Obviously, when the patients were categor-ized according to prior treatment, the frequency of poor

mobilization depended on the degree of previous treatment.In our study, we stopped collecting PBSC in four patients.A previous study showed that the frequency of poormobilizers was 14% in patients mobilized with G-CSFalone,14 which was similar to our experience. The dose-escalation effects of G-CSF alone or in combination withother cytokines in such patients should be investigated toimprove the efcacy of PBSC mobilization.Peripheral blood stem cell yields might be lower in the

G-CSF alone group than in the chemotherapy plus G-CSFmobilization group. However, there is no difference in thenumber of aphereses between the G-CSF alone group andthe chemotherapy plus G-CSF group. A threshold levelof infused progenitors might exist for successful PBSCT.Most pediatric and adolescent patients with solid tumorswho are candidates for high-dose chemotherapy withautologous PBSC support mobilize enough PBSC to ensurethe collection of a PBSC product that is capable of rapidhematological recovery, when mobilization is attemptedwith G-CSF alone or in combination with chemotherapy.Mobilization with G-CSF alone is capable of collectingPBSC over this threshold level, and there were no graftfailures, even in heavily pre-treated patients.In conclusion, mobilization with G-CSF alone was

effective even in pediatric and adolescent patients withsolid tumors who received more than three cycles ofchemotherapy or with extensive irradiation, althoughplatelet recovery following autologous PBSCT was de-layed. Mobilization with G-CSF alone in cancer patientsmight offer some advantages, such as ease of determininga collection schedule without a daily determination ofCD34 cells in the blood, and the avoidance of neutro-penic fever and additional transfusion.

Acknowledgements

We are grateful to Ms A Itho for her technical assistance.

References

1 Handgretinger R, Lang P, Ihm K, Schumn M, Geiselhart A,Koscielniak E et al. Isolation and transplantation ofhighly puried autologous peripheral CD34+ progenitor cells:purging efcacy, hematopoietic reconstitution and long-termoutcome in children with high-risk neuroblastoma. BoneMarrow Transplant 2002; 29: 731736.

2 Arpaci F, Ataergin S, Ozet A, Erler K, Basbozkurt M, OzcanA et al. The feasibility of neoadjuvant high-dose chemotherapyand autologous blood stem cell transplantation in patientswith nonmetastatic high grade localized osteosarcoma. Cancer2005; 104: 10581065.

3 To LB, Haylock DN, Simmons PJ, Juttner CA. The biologyand clinical uses of blood stem cells. Blood 1997; 89: 22332258.

4 Diaz MA, Vincent M, Garcia-Sanchez F, Vicario JL, Madero L.Long-term hematopoietic engraftment after autologous peri-pheral blood progenitor cell transplantation in pediatric patients:effect of the CD34+ cell dose. Vox Sang 2000; 79: 145150.

5 Dazzi C, Cariello A, Rosti G, Argnani M, Sebastiani L,Ferrari E et al. Is there any difference in PBPC mobilizationbetween cyclophosphamide plus G-CSF and G-CSF alone inpatients with non-Hodgkins lymphoma? Leuk Lymphoma2000; 39: 301310.


667


6 Meldgaard-Knudsen L, Jensen L, Gaarsdal E, Nikolaisen K,Johnsen HE. A comparative study of sequential primingand mobilization of progenitor cells with rG-CSF alone andhigh-dose cyclophosphamide plus rG-CSF. Bone MarrowTransplant 2000; 26: 717722.

7 Nakayanasami U, Kanteti R, Morelli J, Klekar A, Al-OlamaA, Keating C et al. Randomized trail of lgrastim versuschemotherapy and lgrastim mobilization of hematopoieticprogenitor cell rescue in autologous transplantation. Blood2001; 98: 20592064.

8 Karanth M, Chakrabarti S, Lovell RA, Harvey C, Holder K,McConkey CC et al. A randomized study comparingperipheral blood progenitor mobilisation using intermediate-dose cyclophosphamide plus lenograstim with lenograstimalone. Bone Marrow Transplant 2004; 34: 399403.

9 Seggewiss R, Buss EC, Herrmann D, Goldschmit H, Ho AD,Fruenhauf S. Kinetics of peripheral blood stem cell mobiliza-tion following G-CSF-supported chemotherapy. Stem Cells2003; 21: 568574.

10 Suh C, Kim S, Kim S, Kim EK, Lee JL, Park KU et al.Initiation of peripheral blood progenitor cell harvest based onperipheral blood hematopoietic progenitor cell counts enume-rated by the Sysmex SE9000. Transfusion 2004; 44: 17621768.

11 Anderlini P, Korbling M. The use of mobilized peripheralblood stem cells from normal donors for allografting. StemCells 1997; 15: 917.

12 Takaue Y, Kawano Y, Abe T, Okamoto Y, Suzue T, ShimizuT et al. Collection and transplantation of peripheral bloodstem cells in very small children weighing 20 kg or less. Blood1995; 86: 372380.

13 Yu J, Leisenring W, Bensinger WI, Holmberg LA, Rowley SD.The predictive value of white cell or CD34+ cell count inperipheral blood for timing apheresis and maximizing yield.Transfusion 1999; 39: 442450.

14 Gojo I, Guo C, Sarkodee-Adoo C, Meisenberg B, Fassas A,Rapoport AP et al. High-dose cyclophosphamide with orwithout etoposide for mobilization of peripheral bloodprogenitor cells in patients with multiple myeloma: efciencyand toxicity. Bone Marrow Transplant 2004; 34: 6976.

15 Kuittinen T, Nousiainen T, Halonen P, Mahlamaki E,Jantunen E. Prediction of mobilization failure in patients withnon-Hodgkins lymphoma. Bone Marrow Transplant 2004; 33:907912.

16 Canales MA, Fernandez-Jimenez MC, Martin A, Arrieta R,Caballero MD, Diez J et al. Identication of factors associatedwith poor peripheral blood progenitor cell mobilization inHodgkins disease. Haematologica 2001; 86: 494498.

17 Ford CD, Green W, Warenski S, Petersen FB. Effects of priorchemotherapy on hematopoietic stem cell mobilization. BoneMarrow Transplant 2004; 33: 901905.

18 Carral A, de la Rubia J, Martin G, Molla S, Martinez J, SanzGF et al. Factors inuencing the collection of peripheral bloodstem cell in patients with acute myeloblastic leukemia andnon-myeloid malignancies. Leuk Res 2003; 27: 512.

19 Koumakis G, Vassiolmanolakis M, Hatzichristou H,Barbounis V, Fillis J, Papanastasiou K et al. Predictivefactors affecting mobilization and peripheral blood stemcell collection using single apheresis for rescuing patientsafter high-dose chemotherapy in various malignancies. BoneMarrow Transplant 1996; 18: 10651072.

20 Noach EJK, Ausema A, van Os R, Akkerman I, Koopal S,Weesing E. Chemotherapy prior to autologous bone marrowtransplantation impairs long-term engraftment. Exp Hemato-logy 2003; 31: 528534.

21 Diaz MA, Villa M, Alegre A, Lamana ML, de la Vega A,Granda A et al. Collection and transplantation of peripheralblood progenitor cells mobilized by G-CSF alone in childrenwith malignancies. Br J Haematol 1996; 94: 148154.


668


peripheral blood stem cell kawano

Documents