Bone Marrow Transplantation (2003) 31, 747–754
& 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00
www.nature.com/bmt
Progenitor Cell Mobilization
G-CSF Alone vs cyclophosphamide plus G-CSF in PBPC mobilization of
patients with lymphoma: results depend on degree of previous pretreatment
G Milone, S Leotta, F Indelicato, S Mercurio, G Moschetti, F Di Raimondo, A Tornello, U Consoli,
G Guido and R Giustolisi
Division of Haematology and Bone Marrow Transplantation, Ospedale Ferrarotto, Catania, Italy
Summary:
We performed a randomized study to compare ‘G-CSF
alone’ (administered at dose of 10 mcg/kg/day) and
‘cyclophosphamide plus G-CSF’ (cyclophosphamide at
dose of 4 g/m2 and G-CSF at dose of 10 lg/kg/day), as
PBPC mobilization schedules in 52 patients with NHL
or HD. Randomization was stratified according to the
amount of previous chemotherapy (p2 and 42 lines of
previous chemotherapy). Mean CD34+ cell peak in P.B.,
mean ‘Total CD34+ cells’ harvested and percentage of
patients successfully mobilized, in the group mobilized
with ‘G-CSF alone’ vs the group mobilized with ‘cyclophosphamide plus G-CSF’, were: 35.3 106 vs 45.8 106/l
(P ¼ 0.3), 5.4 106 vs 6.8 106/kg (P40.9) and 50 vs
61% (P ¼ 0.4). No differences were observed in the
stratum of less pretreated patients. However, in the
stratum of patients who had previously received more than
two lines of chemotherapy, CD34+cell peak (P ¼ 0.05)
and percentage of successful mobilization (P ¼ 0.01) were
higher when ‘cyclophosphamide plus G-CSF’ was used.
Using logistic regression, both age and mobilization with
‘G-CSF alone’ were significantly associated with a low
CD34+ cell peak in P.B. However, in the stratum of less
pretreated patients, only age was significantly associated
with this risk.
Bone Marrow Transplantation (2003) 31, 747–754.
doi:10.1038/sj.bmt.1703912
Keywords: lymphoma; peripheral blood progenitor cell
mobilization; CD34+ cells
High-dose chemotherapy with peripheral blood progenitor
cell (PBPC) rescue has become an accepted option for
treatment of relapsed or poor prognosis Hodgkin disease
(HD) and non-Hodgkin’s lymphomas (NHL).1,2
Although the combination of cyclophosphamide and GCSF is usually used to mobilize PBPC, this scheme is,
however, followed by significant toxicity.3–5 For this reason
many groups are using ‘G-CSF alone’ as mobilization
Correspondence: Dr G Milone, Divisione Clinicizzata di Ematologia,
e Unita’ di Trapianto di Midollo, Ospedale Ferrarotto Via S. Citelli, 6,
95100 Catania, Italy
Received 3 September 2002; accepted 28 October 2002
therapy in patients with lymphomas: a therapy less costly
and not associated with any morbidity.6–9
These two schemes have been compared in breast
cancer 5,10,11 and multiple myeloma patients,4,5 and results
showed a higher mobilization efficiency, in terms of
harvested CD34+ cells, after cyclophosphamide and
G-CSF.
In patients with lymphoma these two schemes have not
always given similar results. In fact, in two studies, the
amount of CD34+ cells harvested after these two regimens
was similar,12,13 while other authors found a higher
mobilizing capacity for the association ‘cyclophosphamide
plus G-CSF’.14–16 Furthermore, when studied as second line
mobilization therapy in patient failing a previous mobilization scheme, ‘G-CSF alone’ and ‘cyclophosphamide plus
G-CSF’ have been considered of equivalent efficacy.17,18
We conducted a prospective randomized study with the
aim of comparing results of ‘cyclophosphamide plus GCSF’ and of ‘G-CSF alone’ as first mobilizing therapy in
patients with Hodgkin’s and non-Hodgkin’s lymphomas.
Since previously administered chemotherapy is an important factor in success of mobilization,9,14,19 we stratified
patients according to the amount of chemotherapy previously
received.
Patients and methods
Elegibility criteria and study design
In this study, we included 52 patients with HD and NHL
who underwent PBPC mobilization treatment in our
Institution from March 1998 to July 2002.
Eligibility criteria were: age less than 70 years, a biopsyproven diagnosis of Hodgkin’s or non-Hodgkin’s lymphoma, normal left ventricular ejection fraction, DLCO greater
than 60% of predicted value, normal renal function with a
creatinine clearance 450 ml/min, bilirubin, aspartate aminotransferase and alanine transferase less than two times
normal, no serological evidence of HIV, HCV or HBV
infection; phase of disease was either first response with
unfavourable prognostic factors or responding relapse. All
patients were on their first mobilization attempt and no
patients on second mobilization were included in this study.
In all cases, a time interval of at least 6 weeks was required
between last chemotherapy and start of mobilization
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
748
Table 1
Patient characteristics
Pts pretreated withp2 lines of
chemotherapy
All patients
Patients number
Mean age (years)
Sex
HD/NHL (%)
Mean interval since last
chemotherapy (months)
Previous B.M. involvement (%)
Low-grade NHL (%)
Mean WBC in P.B. ( 109/l)
G-CSF
CTX+G-CSF
P
G-CSF
26
37.3
62 m 38 f
46/54
3.4
26
43.6
50 m 50 f
27/73
4.3
0.06
0.4
0.14
0.2
19
35
53 m 47 f
48/52
3.0
17
40.2
35 m 65 f
42/58
4.1
42
38
7.2
46
57
5.5
0.7
0.16
0.1
31
31
7.2
30
41
5.8
treatment. During this time interval, criteria of eligibility
were assessed, a complete restaging was performed and
previous chemotherapy received by patients was quantified.
Patients were divided into two categories depending on the
number of lines of chemotherapy that they had received:
patients who had received p2 lines of previous chemotherapy and those who had received 42 lines of previous
chemotherapy. For the purpose of this study, each different
chemotherapy regimen used was counted as one line of
chemotherapy. Chemotherapy regimens were usually administered for number of cycles described in the original
report.
Immediately before the start of mobilization treatment
patients were randomized to receive PBPC mobilization
using ‘G-CSF alone’ or ‘cyclophosphamide plus G-CSF’.
Patient assignment was stratified according to degree of
previous pretreatment (p2 and 42 lines of previous
chemotherapy).
Randomization list was generated, at the start of this
study, using a specifically designed computer program
(RND by S.S.O.S.B., Italy).
Patient characteristics
Patient characteristics studied are reported in Table 1. The
group receiving ‘G-CSF alone’ had a mean age of 37.3
years, while patients receiving ‘G-CSF plus cyclophosphamide’ had a mean age of 43.6 years (P ¼ 0.06). In
stratum of more heavily pretreated patients, mean age of
patients in the group receiving G-CSF was 43.6 and 50
years in the group of patients receiving ‘cyclophosphamide
plus G-CSF’ (P ¼ 0.07). In the stratum of more pretreated
patients, those receiving ‘cyclophosphamide plus G-CSF’
had in higher percentage a diagnosis of NHL with respect
to group receiving ‘G-CSF alone’ (P ¼ 0.02). For all other
variables, no significant differences were found.
First line chemotherapies received by patients at diagnosis were the following disease-oriented regimens: ABVD
(HD), MOPP/ABVD (HD), VACOP-B (NHL), CVP (lowgrade NHL); second line chemotherapies administered were
the following regimens: MINE, IEV, ESHAP and FND.
Mobilization with ‘G-CSF Alone’
Glycosylated-rh-G-CSF (Lenograstim) was employed at
the dose of 10 mcg/kg/day administered in two daily doses
Bone Marrow Transplantation
CTX+G-CSF
Pts pretreated with >2 lines of
chemotherapy
P
G-CSF
CTX+G-CSF
P
0.2
0.2
0.7
0.3
7
43.6
86 m 14 f
43/57
4.4
9
50
78 m 22 f
0/100
4.6
0.07
0.6
0.02
0.6
0.8
0.5
0.4
71
57
7.4
77
88
5.1
0.7
0.14
0.08
by subcutaneous administration from day +1 to the end of
harvesting.
Mobilization with ‘cyclophosphamide plus G-CSF’
Cyclophosphamide was employed at a dose of 4 g/m2
dissolved in 250 ml of 5% dextrose, by i.v. administration
over 1 h on day +1, along with MESNA 3 g/m2 by
continuous infusion for 24 h. From day +4 after cyclophosphamide administration, glycosylated-rh-G-CSF (Lenograstim) was administered at a dose of 10 mg/kg/day in two
daily doses by subcutaneous injection. This therapy was
followed by antiinfectious prophylaxis using fluconazole
and ciprofloxacin.
Collection of PBPC
Collections of PBPC were carried out using a cell separator
(CS 3000 Plus-Baxter); the blood volume processed was
double the calculated blood volume. In the majority of
patients, venous access was obtained by a double lumen
Hickman catheter placed in a central vein. The target
number of CD34+ cells to be obtained in the apheretic
harvests was 5.0 106/kg.
Leukapheresis was started when the threshold value of
CD34+ cells in the peripheral blood was over 20.0 106/l;
patients reaching this threshold were considered ‘successfully mobilized’. Nucleated cells (NC) counts were obtained
by a Sysmex SF-3000 automated cell counter (Sysmex
Corporation, Kobe, Japan).
‘Total CD34+’, ‘Total NC’ and ‘Total CFU-GM’ were
obtained summing the value of the specified determination
in all the aphereses performed in a single mobilization;
‘CD34+ per apheresis’, ‘NC per apheresis’ and ‘CFU-GM
per apheresis’ were obtained dividing ‘Total CD34+’,
‘Total NC’ and ‘Total CFU-GM’ by the numbers of
aphereses performed on each patient.
CD34+ analysis
The CD34+ cell count was determined by flow cytometry.
Briefly, for immunofluorescence analysis, 1.0 106 WBC
from P.B. or from leukapheresis products were incubated
for 15 min at room temperature with phycoerythris (PE)conjugated class III monoclonal anti-CD34 antibody
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
749
Table 2
Results: CD34+ peak in P.B. and percentage of ‘successful mobilization’
All patients
CD34+ Peak in P.B. ( 10 /l)
mean (range)
median
Patients successfully mobilized (%)
G-CSF alone
CTX G-CSF
35.3 (7–159)
19.1
50
45.8 (2.0–151)
39
61
Pts pretreated with
p2 lines of chemotherapy
P
G-CSF alone
CTX G-CSF
0.3
44.3 (8–159)
49
68
48.8 (2.0–151)
33
64
Pts pretreated with
42 lines of chemotherapy
P
G-CSF alone
CTX G-CSF
10.9 (7–16)
8
0
40.2 (2–99)
45
55
P
6
0.4
(Immunotech-Marseille, France) or PE-conjugated isotypic
control mouse IgG1 antibody (Immunotech).
After incubation, erythrocytes were lysed with ammonium chloride buffer for 10 min. The cells were centrifuged
at 600 g for 5 min, washed twice with PBS containing 0.1%
sodium azide, and resuspended in 500 ml of PBS. Immunofluorescence analysis was performed using a FACScalibur
(Becton Dickinson) equipped with argon laser at 488 nm
and 0.3 W. For each sample, 100 000 cells were analysed. In
a forward vs side scatter dot plot, a first region was set up to
separate nucleated cells from debris. A second region
limited by side scatter was set up to limit analysis to the
MNC population. CD34+cells were identified and counted
on another side scatter/PE dot plot, gated in both regions.
For this purpose, CD34+ cells were defined by a highly
fluorescent signal for PE and low side scatter.
The number of CD34+ cells per volume was calculated
by multiplying the CD34 percentage obtained within the
lympho-monocyte gate by leukocyte count obtained from a
haematology analyser and dividing by 100.
0.73
>0.8
0.05
0.01
were compared using the t-test for normally distributed
continuous variables such as age, and Mann–Whitney
U-test for non-normally distributed continuous variables
such as harvested CD34+, CFU-GM and NC. Correlations were studied using the Spearman Rank Test. Risk of
reaching a Peak PB CD34+ value inferior to the median
(23.3 106/l) was determined using logistic regression. To
perform this analysis, all patients were categorized into two
groups according to their P.B. Peak CD34+ values with
respect to the median. This categorical variable was entered
as a dependent variable, whereas age and mobilization type
were entered as independent variables; the reference group
was patients treated with ‘cyclophosphamide plus G-CSF’.
Engraftment times were determined using survival
methods, and results in different groups were compared
using the log-rank test.
A statistics program (Statview, CA, USA) was utilized to
record data and to perform all statistical analyses.
RESULTS
CFU–GM culture
CFU–GM
(granulocyte–macrophage
colony-forming
units) were assayed in harvested PBPC by a culture test
in methocellulose; 1 105 cells were plated in medium
enriched with growth factors (Methocult H04541, Stem
Cell Technologies-Vancouver, Canada). Each sample was
plated in triplicate, cultures were incubated at 371C in a
humidified atmosphere of 5% CO2 and scored under an
inverted microscope after 10 days.
High-dose therapy and PBPC rescue
Patients reaching a minimum harvest of 2.0 106/kg.
CD34+cells received high-dose therapy and PBPC rescue.
In all cases, BEAM was used as the eradicating regimen.20
Patients were nursed in single rooms and received
antibacterial, antifungal and antiviral prophylactic treatment. Neutrophil (N.) engraftment was defined as the first
day in which N. reached a count of at least 0.5 109/l,
while Platelet (PLT) engraftment was defined as the first
day of three consecutive days when the unsupported PLT
count reached the value of 20.0 109/l.
Statistical analysis
Distribution of categorical variables was compared using
the w2 test or Fisher’s exact test. Mean values for groups
Percentage of successful mobilization
The percentage of the total patients which reached a
CD34+ cell peak in peripheral blood superior to 20.
0 106/l, and therefore underwent harvesting of PBPC, was
50% (13/26) with ‘G-CSF alone’ mobilization and 61%
(16/26) with ‘cyclophosphamide plus G-CSF’ (Table 2).
The difference between percentages of mobilizing patients
was not significant (w2 test: P ¼ 0.4).
In patients not heavily pretreated (p2 lines of chemotherapy), the percent successful mobilization was 68%
(13/19) with ‘G-CSF alone’ and 64% (11/17) with
‘cyclophosphamide plus G-CSF’ (w2 test: P4 0.8).
In patients heavily pretreated (42 lines of chemotherapy), the percent successful mobilization was 0% (0/7) with
‘G-CSF alone’ and 55% (5/9) after ‘cyclophosphamide plus
G-CSF’. In patients heavily pretreated, the difference
between percentages of successful mobilization using the
scheme ‘cyclophosphamide plus G-CSF’ compared to ‘GCSF alone’ was significant (P ¼ 0.01).
Peak of CD34+ cells in peripheral blood
Mean values of peak of CD34+ cells in peripheral blood
obtained with ‘G-CSF alone’ and with ‘cyclophosphamide
plus G-CSF’ did not differ significantly: 35.3 106/l. vs
45.8 106/l, respectively (P ¼ 0.3) (Table 2 and Figure 1).
Bone Marrow Transplantation
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
750
Similarly, in patients not heavily pretreated (p2 lines of
chemotherapy) no differences were found. In fact peak
CD34+ cells was 44.3 106/l in patients mobilized with
‘G-CSF alone’ and 48.8 106/l after ‘cyclophosphamide
plus G-CSF’ (P ¼ 0.73).
Conversely, in patients heavily pretreated (42 lines of
chemotherapy) peak CD34+ cells obtained with ‘G-CSF
alone’ was significantly lower than that with ‘cyclophosphamide plus G-CSF’ (10.9 106/l vs 40.2 106/l, P ¼ 0.05).
CD34+ cells, NC and CFU-GM in apheresis products
CD34 + peak Cell/ ×106/I
Results of ‘Total CD34+ cells’, ‘Total NC’, ‘Total CFUGM’ and of ‘CD34+ per apheresis’, ‘NC per apheresis’
and ‘CFU-GM per apheresis’ are reported in Table 3.
180
160
140
120
100
80
60
40
20
0
-20
G-CSF alone
CTX 4 gr/m2
+G-CSF
Factors related to CD34+ cell levels in peripheral blood
Pts pre-treated with
= or < 2 lines of
chemotherapy
Pts pre-treated with
> 2 lines of
chemotherapy
Figure 1
CD34+ peak in P.B. according to degree of pre-treatment
previously received and to scheme of mobilization.
Table 3
‘Total CD34+ cells’ harvested was 5.4 106/kg in
patients mobilized with ‘G-CSF alone’ and 6.8 106/kg in
those mobilized with ‘cyclophosphamide plus G-CSF’
(P40.9).
‘Total NC’ harvested was 5.7 108/kg in the group of
patients mobilized with ‘G-CSF alone’ and 3.6 108/kg, in
patients receiving ‘cyclophosphamide plus G-CSF’, and
this difference was significant (P ¼ 0.008).
‘Total CFU-GM’ harvested was 3.5 105/kg in patients
mobilized with ‘G-CSF alone’ and 4.1 105/kg in patients
mobilized with ‘cyclophosphamide plus G-CSF’, difference
was not significant (P ¼ 0.8).
No significant differences were detected in either ‘Total
CD34+’ cells or in ‘CD34+ per apheresis’, when the
stratum of patients pretreated with r2 lines of chemotherapy was considered. In the group of patients pretreated
with 42 lines of chemotherapy, none had a successful
mobilization using ‘G-CSF alone’.
Mean number of aphaereses was 1.7, 1.6 and 2.0
for, respectively, the whole population of patients, patients
not heavily pretreated and those who were heavily
pretreated.
In the whole population of patients, a negative correlation
(Figure 2) was observed between age and mean peak of
CD34+ cells in peripheral blood (Spearman Rank
correlation test, r ¼ 0.35, P ¼ 0.011).
No significant correlation was observed for time interval
from last chemotherapy to mobilization and CD34+ cells
peak in peripheral blood (Spearman Rank correlation test,
r ¼ 0.02, P ¼ 0.88).
Results: total CD34+, total NC, total CFU-GM, CD34+/apheresis, NC/apheresis and CFU-GM/apheresis
All patients
Total CD34+ 10 /kg
mean
(range)
median
Total NC 108/kg
mean
(range)
median
Total CFU-GM 105/kg
mean
(range)
median
CD34+ 106/kg/apheresis
mean
(range)
median
NC 108/kg/apheresis
mean
(range)
median
CFU-GM 105/kg/apheresis
mean
(range)
median
G-CSF alone
CTX+ G-CSF
5.4
(1.5–15)
5.4
6.8
(2.7–24)
4.6
5.7
(2.1–11.8)
5.0
3.6
(1.0–7)
3.4
3.57
(0.5–11)
3.1
Pts pretreated with
p2 lines of chemotherapy
P
G-CSF alone
CTX+ G-CSF
5.4
(1.5–15)
5.4
5.69
(2.7–22)
4.4
0.008
5.7
(2.1–11.8)
5.0
4.1
(0.3–19)
2.0
0.8
3.8
(1.3–15)
2.7
4.6
(0.9–22)
3.3
3.9
(1.0–7.6)
3.9
2.5
(0.2–11)
1.5
Pts pretreated with
>2 lines of chemotherapy
P
G-CSF alone
CTX+ G-CSF
0.4
—
9.4
(3.7–24)
6.0
3.4
(1.0–5.0)
3.4
0.007
—
4.2
(2.0–7.0)
3.5
3.5
(0.5–11)
3.1
4.0
(0.3–19)
1.5
0.7
—
4.36
(0.8–10.5)
3.0
0.7
3.8
(1.3–15)
2.7
4.6
(0.9–22)
2.3
0.8
—
4.52
(1.8–8.0)
3.9
2.2
(1.0–3.5)
2.3
0.01
3.9
(1.0–7.6)
3.9
2.2
(1.0–3.5)
2.4
0.02
—
2.2
(1.0–3.5)
2.3
3.3
(0.1–19)
1.39
0.8
2.5
(0.2–11)
1.5
3.8
(0.1–19)
1.11
>0.9
—
2.29
(0.8–5.0)
1.5
6
>0.9
In the group of patients pretreated with >2 lines of chemotherapy none had a successful mobilization using ‘G-CSF alone’.
Bone Marrow Transplantation
CD34 + peak In P.B. Cell/ ×106/I
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
180
160
140
120
100
80
60
40
20
0
-20
10
20
30
40
50
Age
60
70
80
Figure 2 Age and CD34+ peak in P.B.
In the whole population of patients, no differences were
observed in peripheral blood CD34+ cells peaks for
groups identified on the basis of sex (P ¼ 0.5), NHL or
HD diagnosis (P ¼ 0.08), low grade-NHL diagnosis
(P ¼ 0.1) and previous bone marrow involvement (P ¼ 0.6).
Multivariate analysis of the risk for a low CD34 peak
Since we found that age was significantly correlated with
mobilization efficiency and since mean age was different in
the two mobilization groups, we used multivariate analysis
to study the risk of a low CD34+ cell peak associated,
independently, with mobilization schemes and with age.
Logistic regression showed that age (odds ratio: 1.081;
Wald test: P ¼ 0.008; likelihood ratio test: P ¼ 0.05) and
mobilization using ‘G-CSF alone’ (odds ratio: 3.383; Wald
test: P ¼ 0.06; likelihood ratio test: P ¼ 0.05) increased,
independently, the risk of obtaining a PB CD34+ peak
below the median (23.3 106/l).
However, when analysis was computed in the stratum of
less heavily pretreated patients, only age was found to
increase the risk of having a peak of CD34+ cells lower
than the median value (odds ratio: 1.062; Wald test:
P ¼ 0.05; likelihood ratio test: P ¼ 0.03), while no effect was
attributed to ‘G-CSF alone’ (Wald test: P ¼ 0.6; likelihood
ratio test: P ¼ 0.6).
Logistic regression was not computed in the more heavily
pretreated group because of the small number of patients in
this stratum.
G-CSF was employed post-transplantation according to
an institutional randomized trial and it was administered in
5/8 patients who received PBPC mobilized with ‘G-CSF
alone’ and in 3/8 patients who received PBPC mobilized
with ‘cyclophosphamide plus G-CSF’.
Median time to reach a neutrophil count 40.5 109/l
was 11.2 days (range 9–15) in the ‘G-CSF alone’ mobilized
group and 11.8 days (range 10–16) in the ‘cyclophosphamide plus G-CSF’ mobilized group (log-rank test:
P-value ¼ 0.5).
Median time to reach a platelet count420 109/l was
12.2 days (range 10–14) in the ‘G-CSF alone’ mobilized
group and 14.5 (range 11–23) days in the ‘cyclophosphamide plus G-CSF’ mobilized group (log-rank test:
P-value ¼ 0.33).
751
Possibility of obtaining a stem cell harvest in patients
failing first mobilization
In total, 16 patients were successfully mobilized using
‘cyclophosphamide plus G-CSF’, while 10 patients did not
experience a successful mobilization, seven patients received a second course of mobilization using chemotherapy
(IEV, VP-16 as single agent or Cyclophosphamide 4 g/m2)
plus G-CSF, and 2/7 had a successful apheretic harvest.
Thus, five patients out of seven also failed this second
course of mobilization, and 4/5 underwent successful bone
marrow harvesting. Overall, in the group receiving ‘cyclophosphamide plus G-CSF’ as first mobilization treatment,
an adequate stem cell inoculum was, eventually, obtained in
22/26 patients (84%).
Overall, 13 patients were successfully mobilized using
G-CSF alone, while 13 patients failed the first course of
mobilization. In this group, 10 patients received a second
course of mobilization using cyclophosphamide 4 g/m2 plus
G-CSF, and 4/10 had a successful apheretic harvest. Of the
six patients who also failed the second mobilization course,
four had a successful bone marrow harvest.
Overall, in the group receiving ‘G-CSF alone’ as first
mobilization treatment, adequate stem cells for rescue were
obtained in 21/26 patients (80%).
The difference in the overall percentage of patients
reaching an adequate stem cell inoculum between groups in
study was not significant (w2 test: P-value ¼ 0.7).
Discussion
Transplantation results in successfully mobilized patients
A total of 29/52 patients were successfully mobilized using
one of the two regimens considered as first mobilization
treatment and, from November 1998 to March 2002, 16/29
patients have already received high-dose therapy with
PBPC rescue. All transplanted patients had engraftment
and no regimen-related mortality was documented. Eight
patients have been transplanted using G-CSF mobilized
PBPC and eight patients using ‘cyclophosphamide plus
G-CSF’ mobilized PBPC. The first group received a mean
of 6.9 106/kg CD34+ (range 4.8–11.0), while the second
received a mean of 6.3 106/kg CD34+ (range 2.5–22.0),
(Mann–Whitney U-test, P-value ¼ 0.02). Prophylactic
Peaks in P.B. of CD34+cells, percentage of successfully
mobilized patients and contents of apheretic harvests did
not differ, on univariate analysis, between the two groups
treated with ‘G-CSF alone’ and with ‘cyclophosphamide
plus G-CSF’.
However, in our study we also found, as recently
reported by Gazitt et al,19 that age is an important factor
for efficiency of mobilization. We therefore used multivariate analysis to determine the efficiency of mobilization
of ‘G-CSF alone’ and of ‘cyclophosphamide plus G-CSF’,
taking into account the effect of age. Results showed that
mobilization with ‘G-CSF alone’ has a higher probability
of a CD34+ peak below the median. According to this
Bone Marrow Transplantation
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
752
analysis, ‘cyclophosphamide plus G-CSF’ has, indeed,
higher mobilization efficiency than ‘G-CSF alone’, in
accordance with reports by other authors.14–16 A higher
number of CD34+ cells after ‘cyclophosphamide plus
G-CSF’ has been found by Meldgaard Knudsen et al,16
the crossover design of this study allowed a high sensitivity
in detecting differences between the two mobilization
schemes.
A high CD34 + cell content in apheretic harvests was
also found using ‘cyclophosphamide 5 g/m2 plus G-CSF’
with respect to ‘G-CSF alone’ in a recently published
study.15 Some ways in which this study differs from
ours should be pointed out: cyclophosphamide dose
was higher, the threshold of CD34+ cells in P.B. used to
start the apheresis was lower and the volume of blood
processed was greater. The latter two points are the most
likely explanations for the higher percentage of patients
successfully harvested reported for both mobilization
arms.
However, no differences in CD34+ cell contents of
apheretic harvest after mobilization using ‘cyclophosphamide plus G-CSF’ and ‘G-CSF alone’ were reported by
Cesana et al.13 The relatively small size of this study may be
the reason for failure to detect differences in the two
mobilization schemes.
In the study of Kroger et al,12 the same results of
harvested CD34+ cells were obtained in lymphoma
patients receiving chemotherapy plus G-CSF or ‘G-CSF
alone’. The therapy administered differed in that a very
high dose of G-CSF was employed. Furthermore, the
schedule of chemotherapy used included BCNU, a drug
toxic to the haematopoietic stem cell,9 which may, therefore, have impaired the process of mobilization in the arm
where chemotherapy and G-CSF were combined.
No data are available in the literature comparing
‘cyclophosphamide plus G-CSF’ and ‘G-CSF alone’ in
patients stratified according to degree of chemotherapy
pretreatment.
In this study of patients with lymphoma who were not
heavily pretreated, ‘G-‘CSF alone’ gave, on univariate
analysis, results comparable to ‘cyclophosphamide plus
G-CSF’. Multivariate analysis confirmed that, in this
stratum of nonheavily pretreated patients, mobilization
using ‘G-CSF alone’ did not increase the risk of a CD34+
cell peak in the P.B. being below the median, and therefore
of an unsuccessful mobilization, even when the effect of age
was taken into account.
Conversely, a higher mobilization efficiency with the
association ‘cyclophosphamide plus G-CSF’ was evident
on univariate analysis, in both peak of CD34+ cells and
percentage of successfully mobilized patients when the
stratum of more heavily pretreated patients was examined.
In fact, in this stratum these parameters were lower using
‘G-CSF alone’ compared to ‘cyclophosphamide plus
G-CSF’.
The lower mobilization efficiency of ‘G-CSF alone’ in the
stratum of more heavily pretreated patients may have practical
implications in the choice of a mobilizing scheme. In heavily
pretreated patients, the use of ‘cyclophosphamide plus
G-CSF’ may be preferable since this yields a substantially
higher probability of successfully mobilization.
Bone Marrow Transplantation
However, apart from the number of CD34+ cells
harvested, other issues should be considered in evaluating
a mobilization scheme.
First, although the number of CD34+ cells is recognized
as important in evaluating engraftment potential after
autologous PBPC transplants, CD34+ cells are nevertheless heterogeneous, and some subsets of CD34+ cells
could be more important than others. CD34+ cells
obtained from PB after different mobilization schemes
show differences in surface antigens 12,21 which could imply
different CD34+ subpopulation distributions and hence
different engraftment potentials. In this regard it is
important to note that in studies comparing ‘cyclophosphamide plus G-CSF’ and ‘G-CSF alone’, different
amounts of CD34+ cells were obtained even though
engraftment times were the same with both mobilizing
schemes.4,15 Our own clinical experience, although limited
because of the smaller number of patients who have been
transplanted to date using PBPC harvested during a first
mobilization, is in accordance with these results.
Besides mobilizing efficiency and engraftment potential,
other factors should be taken into consideration as
clinically important. One of these is neoplastic contamination of PBPC. The importance of neoplastic contamination
of infused stem cells in patients with lymphomas is still an
unresolved issue. The usefulness of purging has been
demonstrated in patients with follicular lymphomas,22
while the importance of neoplastic contamination in
patients with other types of lymphoma has not been so
clearly shown.23 Lymphoma cells circulate in P.B. in many
patients,24,25 and they can be mobilized by current
mobilization strategies;24 however, the seeding efficiency
of these cells and hence their role in determining relapse
after autologous transplantation is not clear.26,27 Tumour
cell contamination of apheretic harvests obtained after
‘cyclophosphamide plus G-CSF’ or after ‘G-CSF alone’ has
been studied in patients with breast cancer and in patients
with lymphomas and no differences have been detected.11,15,28
‘G-CSF alone’, compared to ‘cyclophosphamide plus GCSF’, results in apheretic harvests with a higher content of
nucleated cells and lymphocytes; furthermore, a fast immune
recovery has been reported after autologous transplantation
of PBPC harvested using ‘G-CSF alone’.29 Such a rapid
immune system recovery could have a beneficial effect on
long-term results of high-dose therapy and autologous
transplantations. In fact, it has recently been reported that
after autologous PBPC transplants, an early immune system
recovery is associated with a better outcome.30
Finally, it should be considered that in our study many
patients failing the first mobilization treatment were
‘rescued’ by a second mobilization course or by bone
marrow harvesting. Thus, overall, comparable results were
obtained in terms of the number of patients who could
eventually have an adequate inoculum and could receive
high-dose therapy, and this regardless of which of the two
mobilization schemes was first employed.
We conclude, therefore, that ‘cyclophosphamide plus GCSF’ has, in the population of patients studied as a whole,
statistically higher mobilization efficiency. This superiority
is however limited to the stratum of patients who had
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
previously received heavier chemotherapy treatment. In
such patients, therefore, the combination of ‘cyclophosphamide plus G-CSF’ may be preferable, given the
significantly higher probability of successful mobilization.
In less heavily pretreated patients, no mobilizing
advantage of ‘cyclophosphamide plus G-CSF’ was apparent in our study, although it is possible that this advantage
may be detectable if a larger number of patients is
studied.
Nevertheless, since ‘G-CSF alone’ shows good mobilization efficacy with no evidence of a higher risk of neoplastic
contamination, and offers the advantages of low cost, low
morbidity and fast immune recovery, in this patient subset
it can be considered an adequate mobilization approach.
Further comparison between these two mobilization
regimens must be undertaken in wider studies so that any
possible differences in survival offered by these two
schemes may be fully apparent.
In the area of HSC mobilization there is need for further
studies directed at ensuring an improvement in the
harvested PBPC product for as many patients as possible,
and at achieving this goal without the toxicity of
chemotherapy.31
753
8
9
10
11
12
13
Acknowledgements
We are grateful to Mike Wilkinson for the English translation of
the manuscript.
14
References
15
1 To LB, Roberts MM, Haylock DN et al. Comparison of
haematological recovery times and supportive care requirements of autologous recovery phase peripheral blood stem cell
transplants, autologous bone marrow transplants and allogeneic bone marrow transplants. Bone Marrow Transplant
1992; 9: 277–284.
2 Schmitz N, Linch DC, Dreger P et al. Randomized trial of
filgrastim-mobilized peripheral blood progenitor cell transplantation versus autologous bone-marrow transplantation in
lymphoma patients. Lancet 1996; 347: 353–357.
3 Alegre A, Tomas JF, Martinez-Chomorro C et al. Comparison
of peripheral blood progenitor cell mobilization in patients
with multiple myeloma: high dose cyclophosphamide plus
GM-CSF vs G-CSF alone. Bone Marrow Transplant 1997; 20:
211–217.
4 Desikan KR, Barlogie B, Jagannath S et al. Comparable
engraftment kinetics following peripheral blood stem cell
infusion mobilized with granulocyte colony stimulating factor
with or without cyclophosphamide in multiple myeloma. J Clin
Oncol 1998; 6: 1547–1553.
5 Akard LP, Thompson JM, Dugan MJ et al. Matched pair
analysis of hemopoietic progenitor cell mobilisation using GCSF vs cyclophosphamyde, etoposide and G-CSF in enhanced
CD34+ cell collections are not necessarily cost-effective. Biol
Blood Marrow Transplant 1999; 5: 379–385.
6 Kroger N, Zeiter W, Hassan HT et al. Successful mobilisation
of peripheral blood stem cells in heavily pretreated myeloma
patients with G-CSF alone. Ann Hematol 1998; 76: 257–262.
7 Grimaz S, Damiani D, Michieli M et al. Mobilisation of
peripheral blood progenitor cells with G-CSF alone in
lymphoma patients: steady state circulating progenitor cells
16
17
18
19
20
21
22
count predicts the autograft yield. Haematologica 2000; 85:
102–103.
Feremans SW, Le Moine F, Ravolet C et al. Optimal
blood stem cell mobilisation using 10 micrograms/kg
granulocyte colony-stimulating factor (G-CSF) alone for
high-dose melphalan intensification in multiple myeloma:
an intrapatient controlled study. Am J Hematol 1994; 47:
135–138.
Canales MA, Fernandez-Jimenez MC, Martin A et al.
Identification of factors associated with poor peripheral blood
progenitor cell mobilization in Hodgkin disease. Haematologica 2001; 86: 4949–4958.
Meisenberg B, Brhem T, Schmekel A et al. A combination of
low-dose cyclophosphamide and colony stimulating factors is
more cost-effective than granulocyte-colony stimulating factors alone in mobilizing peripheral blood stem and progenitor
cells. Transfusion 1998; 38: 209–215.
Bertolini F, Lanza A, Peccatori F et al. Hemopoietic
progenitor cell collection and neoplastic cell contamination
in breast cancer patient receiving chemotherapy plus granulocyte-colony stimulating factor (G-CSF) or G-CSF alone for
mobilisation. Ann Oncol 1998; 9: 913–916.
Kroger N, Zeiler W, Fehese N et al. Mobilizing peripheral
blood stem cell with high-dose G-CSF alone is effective as
with dexa-BEAM plus G-CSF in lymphoma patients. Br J
Haematol 1998; 102: 1101–1106.
Cesana C, Carlo-Stella C, Regazzi E et al. CD34 cells
mobilized by cyclophosphamide and granulocyte colonystimulating factor (G-CSF) are functionally different from
CD34 cells mobilized by G-CSF. Bone Marrow Transplant
1998; 21: 561–568.
Moskowitz CH, Glassman JR, Wuest D et al. Factors affecting
mobilisation of peripheral blood progenitor cells in patients
with lymphoma. Clin Cancer Res 1998; 4: 311–316.
Narayanasami U, Kanteli R, Morelli J et al. Randomized trial
of filgrastim versus chemotherapy and filgrastim mobilization
of hematopoietic progenitor cells for rescue in autologous
transplantation. Blood 2001; 98: 2059–2064.
Meldgaard Knudsen L, Jensen L, Gaarsdal E et al.
A comparative study of sequential priming and mobilisation
of progenitor cells with rh-G-CSF alone and high dose cyclophosphamide plus rh-G-CSF. Bone Marrow Transplant 2000;
26: 717–722.
Fraipont V, Sautois B, Baudoux E et al. Successful mobilization of peripheral blood HPCs with G-CSF alone in patients
failing to achieve sufficient numbers of CD34+ cells and/or
CFU-GM with chemotherapy and G-CSF. Transfusion 2000;
40: 339–347.
Weaver CH, Tauer K, Zhen B et al. Second attempts at
mobilization of peripheral blood stem cell in patients
with initial low CD34+ cell yields. J Hematother 1998; 7:
241–249.
Gazitt Y, Shaughnessy P, Liu Q. Differential mobilization
of CD34+ cells and lymphoma cells in non-Hodgkin’s
lymphoma patients mobilized with different growth factors.
J Hematother Stem Cell Res 2001; 10: 167–176.
Wills W, Chopra R, McMillan A et al. BEAM chemotherapy
and autologous bone marrow transplantation for patients with
relapsed or refractory Non-Hodgkin’s lymphoma. J Clin Oncol
1995; 13: 588–595.
To LB, Hylock DN. A comparative study of the phenotype
and proliferative capacity of peripheral blood (PB) CD34+
cells mobilized by four different protocols and those of steady
phase PB and bone marrow CD34+ cells. Blood 1994; 84:
2930–2939.
Gribben JG, Freedman AS, Neuberg D et al. Immunologic
purging of marrow assessed by PCR before autologous bone
Bone Marrow Transplantation
G-CSF alone vs cyclophosphamide plus G-CSF
G Milone et al
754
23
24
25
26
marrow transplantation for B-cell Lymphoma. N Engl J Med
1991; 325: 1525–1535.
Bierman PJ, Sweetenauham T, Loberizza F et al. Syngenic
hematopoietic stem cell transplantation for Non Hodgkin’s
Lymphoma (NHL): comparison with allogenic and autologous
transplants suggests a role for purging. Proc Am Soc Clin
Oncol 2001; 20: 1(abstract 15).
Jaquy C, Soree A, Lambert F et al. A quantitative study of
peripheral blood stem cell contamination in diffuse large-cell
non-Hodgkin’s lymphoma: one-half of patients significantly
mobilize malignant cells. Br J Haematol 2000; 110: 631–637.
Gribben JG, Freedman AS, Woo SD et al. All advanced stage
non-Hodgkin lymphomas with polymerase chain reaction
amplifiable breakpoint of bcl-2 have residual cells containing
the bcl-2 rearrangement at evaluation and after treatment.
Blood 1991; 78: 3275–3280.
Gidali J, Feher I. Lymphoma cell contamination of PBPC: a
murine model. Bone Marrow Trasplant 1999; 24: 617–620.
Bone Marrow Transplantation
27 Demirkazik A, Kessinger A, Armitage JO et al. Progenitor and
lymphoma cells in blood stem cell harvests: impact on survival
following transplantation. Bone Marrow Transplant 2001; 28:
207–212.
28 Passos-Coelho JL, Ross AA, Kahn DJ et al. Similar breast
cancer cell contamination of single-day peripheral blood
progenitor cell collections obtained after priming with
hemopoietic growth factor alone or after cyclophosphamide
followed by growth factor. J Clin Oncol 1996; 14: 2569–2575.
29 Rosillo MC, Ortuno F, Moraleda JM et al. Immune recovery
after autologous or rhG-CSF primed PBPC transplantation.
Eur J Haematol 1996; 56: 301–307.
30 Porrata LF, Gertz MA, Inwards DJ et al. Early lymphocyte
recovery predicts superior survival after autologous hematopoietic stem cell transplantation in multiple myeloma or Non
Hodgkin’s lymphoma. Blood 2001; 98: 579–585.
31 Akard L. Optimum methods to mobilize stem cells. J Clin
Oncol 2000; 18: 3063.