Download Autologous Bone Marrow Stem Cell Mobilization Induced by

Journal of the American College of Cardiology
© 2006 by the American College of Cardiology Foundation
Published by Elsevier Inc.
Vol. 48, No. 8, 2006
ISSN 0735-1097/06/$32.00
doi:10.1016/j.jacc.2006.07.044
EXPEDITED REVIEW
Autologous Bone Marrow Stem Cell Mobilization
Induced by Granulocyte Colony-Stimulating Factor
After Subacute ST-Segment Elevation Myocardial
Infarction Undergoing Late Revascularization
Final Results From the G-CSF-STEMI (Granulocyte Colony-Stimulating
Factor ST-Segment Elevation Myocardial Infarction) Trial
Markus G. Engelmann, MD,* Hans D. Theiss, MD,* Christine Hennig-Theiss,* Armin Huber, MD,†
Bernd J. Wintersperger, MD,† Anja-Eva Werle-Ruedinger,† Stefan O. Schoenberg, MD,†
Gerhard Steinbeck, MD,* Wolfgang-M. Franz, MD*
Munich, Germany
The purpose of this investigator-driven, prospective, randomized, double-blinded, placebocontrolled phase II study was to compare the effects of granulocyte colony-stimulating factor
(G-CSF) on the improvement of myocardial function in patients undergoing delayed percutaneous coronary intervention (PCI) for ST-segment elevation myocardial infarction (STEMI).
BACKGROUND Experimental and early clinical studies suggest that transplantation of stem cells improves
cardiac regeneration and neovascularization after acute myocardial infarction. Most investigators have utilized either a direct injection or intracoronary infusion of bone marrow– derived
cells, but early cytokine-mediated mobilization of stem cells has been reported to show similar
improvement in cardiac function.
METHODS
Forty-four patients with late revascularized subacute STEMI were treated either with G-CSF
or placebo over 5 days after successful PCI. Primary end points were change of global and
regional myocardial function from baseline (1 week after PCI) to 3 months after PCI assessed
by magnetic resonance imaging (MRI). Secondary end points consisted of characterization of
mobilized stem cell populations, assessment of safety parameters up to 12 months including
6-month angiography, as well as myocardial perfusion assessed by MRI.
RESULTS
Global myocardial function from baseline (1 week after PCI) to 3 months improved in both
groups, but G-CSF was not superior to placebo (⌬ejection fraction 6.2 ⫾ 9.0 vs. 5.3 ⫾ 9.8%, p
⫽ 0.77). A slight but non-significant improvement of regional function occurred in both
groups. Granulocyte colony-stimulating factor resulted in mobilization of endothelial
progenitor cell populations and was well tolerated with a similar rate of target lesion
revascularization from in-stent restenosis. In both groups major adverse cardiovascular events
occurred in a comparable frequency. Granulocyte colony-stimulating factor resulted in
significant improvement of myocardial perfusion 1 week and 1 month after PCI.
CONCLUSIONS Granulocyte colony-stimulating factor treatment after PCI in subacute STEMI is feasible
and relatively safe. However, patients do not benefit from G-CSF when PCI is performed
late. Granulocyte colony-stimulating factor results in improved myocardial perfusion of the
infarcted area, which may reflect enhanced neovascularization. (J Am Coll Cardiol 2006;48:
1712–21) © 2006 by the American College of Cardiology Foundation
OBJECTIVES
Late coronary reperfusion is frequently associated with left
ventricular remodeling leading to sudden cardiac death or
progressive heart failure. The outcome of patients suffering
from subacute myocardial infarction (MI) is considered to be
serious; patients are threatened by progressive myocardial
From the *Medical Clinic I—Department of Cardiology, †Department of Clinical
Diagnostic Radiology, Ludwig Maximilians University, Klinikum Grosshadern,
Munich, Germany. This study was supported by research grants from Amgen GmbH,
Munich, Germany; E. Lilly Deutschland GmbH, Bad Homburg, Germany; and
Altana, Konstanz, Germany. The Ludwig Maximilians University is the holder of a
pending patent (“Uses and methods for treating ischemia,” EP 03 02 4526.0 and US
60/514,474) claiming a second medical use of G-CSF to treat ischemic organ failure.
Elements of this study are part of the theses of C.H.T. and A.E.W.R.
Manuscript received January 20, 2006; revised manuscript received June 29, 2006,
accepted July 3, 2006.
dysfunction and increased mortality resulting from the long
time interval between onset of infarction to revascularization
(1). Animal experiments have shown that application of
See page 1722
granulocyte colony-stimulating factor (G-CSF) after MI
can improve mortality and ameliorate myocardial damage
(2). Early non-placebo controlled human trials report
safety of G-CSF administration after immediate percutaneous coronary intervention (PCI) and improvement of
global left ventricular function (3,4). However, the use of
G-CSF was made uncertain by a report of increased rate
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Abbreviations and Acronyms
EF
⫽ ejection fraction
FIRSTLINE-AMI ⫽ Front-Integrated Revascularization
and Stem Cell Liberation in
Evolving Acute Myocardial
Infarction by Granulocyte ColonyStimulating Factor Trial
G-CSF
⫽ granulocyte colony-stimulating
factor
ISR
⫽ in-stent restenosis
MACE
⫽ major adverse cardiovascular
events
MI
⫽ myocardial infarction
MRI
⫽ magnetic resonance imaging
PCI
⫽ percutaneous coronary
intervention
STEMI
⫽ ST-segment elevation myocardial
infarction
TLR
⫽ target lesion revascularization
of in-stent restenosis (ISR) when G-CSF was administered before PCI (5).
The aim of this investigator-driven, prospective, randomized, double-blinded clinical study was to investigate safety
and efficacy of G-CSF in patients undergoing delayed
revascularization after subacute ST-segment elevation myocardial infarction (STEMI).
MATERIALS AND METHODS
Study protocol. Starting in December 2002, we prospectively enrolled patients suffering from subacute STEMI
with late revascularization achieved by PCI. Inclusion criteria were subacute STEMI, onset of pain more than 6 h
and up to 7 days, akinesia of at least 1 myocardial segment
demonstrated by echocardiography at admission, suitable
for PCI of the infarct-related artery, and no contraindications against electrocardiogram (ECG)-triggered magnetic
resonance imaging (MRI) (Fig. 1). Patients who were
clinically unstable, had other severe underlying illnesses, or
contraindications against G-CSF were excluded. Patients
having aspirin intolerance or currently receiving steroids,
immunosuppressants, or cytostatics were also excluded.
None of the patients had a history of MI.
After acute PCI of the infarct-related artery using bare
metal stents, patients were treated with clopidogrel for at
least 4 weeks. Patients were randomized to receive either
G-CSF (Filgrastim, Amgen GmbH, Munich, Germany) at
a dose of 10 ␮g/kg body weight/day subcutaneously or
placebo (saline). Randomization and preparation of the
study medication using neutral syringes were independently
performed by the department of pharmacy of our institution. Follow-up visits were performed at 1 and 3 months
and included clinical status, laboratory examinations, ECG,
echocardiography, safety, adverse events, and medications.
All patients received aspirin, clopidogrel, angiotensinconverting enzyme inhibitors, beta-blockers, and statins at
Engelmann et al.
G-CSF in Myocardial Infarction
1713
discharge. The use of clopidogrel was mandatory for 4
weeks after PCI, and a recommendation to the general
practitioner/cardiologist was given by our institution regarding continuation of clopidogrel for 3 to 6 months. After
completion of the 3-month follow-up visit, patients entered
an observational study in order to assess long-term safety
and clinical outcome of treatment up to 1 year. To assess
occurrence of ISR, coronary angiography was performed 6
months after enrollment (Fig. 1).
The study was conducted according to the national and
international regulations and was approved by the university’s ethics committee. All patients gave their written informed consent.
Study objectives. PRIMARY END POINTS. The primary end
points were changes of global and regional cardiac function
using MRI from baseline (1 week after PCI) to 3 months of
follow-up. Global function was determined by left ventricular ejection fraction (EF). Regional myocardial function
was assessed as segmental systolic wall thickening of the
infarct area.
SECONDARY END POINTS. Secondary end points comprised
changes of end-diastolic and end-systolic myocardial thickness, end-diastolic volume, end-systolic volume, and infarct
volume using MRI from baseline (1 week after PCI) to 3
months of follow-up. Furthermore, change of myocardial
perfusion parameters were assessed from baseline to 3
months. Occurrence of major adverse cardiac events
(MACE), such as death, repeat MI or acute coronary
syndromes, coronary artery bypass grafting, and reintervention, as well as spontaneously reported adverse events were
followed up to 1 year after PCI. Changes of blood count and
liver enzymes, number and characterization of mobilized
stem cells (CD34⫹/c-kit⫹, CD34⫹/CD31⫹, CD34⫹/
CD133⫹), and changes of inflammatory parameters were
analyzed.
The sample size was determined to assess efficacy of
G-CSF with regard to improvement of segmental systolic
wall thickening and left ventricular EF. Detection of a
difference of 1 mm in systolic wall thickening, and 8% in left
ventricular EF, respectively, with an 80% power and an ␣
error of 5%, would require 36 patients (18 patients for each
treatment group). We adjusted the sample size for an
estimated 10% loss of follow-up, which resulted in 20
patients in each group and a total sample size of 40 patients.
Laboratory analyses. Complete blood count was routinely
assessed using automated laboratory cell counter. C-reactive
protein analysis was performed using turbidometry. Cytokine level of interleukin-6 was assessed using enzyme-linked
immunoadsorbent assay. Liver enzymes measurement was
performed in a standard automated analyzer.
Flow cytometry. Cytometric analysis was performed using
a flow cytometer (FACScan, Becton Dickinson, Heidelberg, Germany). Each analysis included 100,000 events. For
immunophenotyping, we used the monoclonal antibodies
against CD31, CD34, CD45, CD117 (c-kit), CD133,
1714
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Figure 1. Study design. F/U ⫽ follow-up; G-CSF ⫽ granulocyte colony-stimulating factor; MRI ⫽ magnetic resonance imaging; PCI ⫽ percutaneous
coronary intervention; s.c. ⫽ subcutaneously; STEMI ⫽ ST-segment elevation myocardial infarction.
CXCR4 (Clone 12G5, R&D Systems, Minneapolis, Minnesota), conjugated with fluorescein isothiocyanate, phycoerythrin, or phycoerythrin cyanine-5 (BD PharMingen/
Coulter Immunotech, Hamburg, Germany).
MRI. The MRI examinations were carried out on a
1.5-T whole-body scanner (Magnetom Sonata, Siemens
Medical Solutions, Erlangen, Germany) at 3 time points;
the first examination was performed 1 week after successful PCI (baseline, 1 week after PCI) to avoid overestimation of myocardial impairment resulting from stunning (6). Follow-up MRI was performed at 1 and 3
months. Cardiac functional imaging was based on a
segmented CINE TrueFISP pulse sequence using a
shared echo technique (temporal resolution: 42 ms, voxel
size: in-plane resolution 1.3 ⫻ 1.5 mm2, slice thickness:
8 mm). Functional assessment was performed with a
stack of slices with 1-cm distance in a double oblique
short-axis orientation (7). Regional systolic myocardial
thickening was assessed on a segmental base (16segement model) as absolute and relative values determined in areas of infarction, border zone, and remote
areas not affected by infarction. Myocardial segments that
were related to the infarct artery and presented with loss
of function as well as demonstrated late enhancement
were judged to be infarct segments. Border zone segments were defined as the 2 segments adjacent to the
infarct region located in the same slice; diastolic and
end-systolic thickness were added and divided by 2.
Remote segments were those related to a non-infarct
artery, had preserved function, and did not show late
enhancement. The myocardial viability was assessed using inversion recovery T1w contrast-enhanced MRI using the late enhancement technique using gadobenate
dimeglumine (Multihance, Altana, Konstanz, Germany, 0.1
mmol/kg bodyweight), as previously described (8).
Myocardial perfusion was assessed using a T1-weighted
saturation recovery gradient echo sequence with prospective
ECG triggering during the first path of contrast agent.
Three slices with a thickness of 10 mm were acquired in a
basal, mid-papillary, and apical position in a short-axis view
every heart beat while breath holding (field-of-view 340 ⫻
265 mm2, in-plane spatial resolution 2.7 ⫻ 2.1 mm).
Hyperemia was induced with a continuous intravenous
infusion of 140 ␮g/kg · min⫺1 adenosine (Adenoscan,
Sanofi, Munich, Germany, bolus 0.05 mmol/kg gadobenate
dimeglumine, flow rate 5 ml/s). For perfusion analysis, the
left ventricular myocardium was divided into 6 equiangular
segments per slice, and signal intensity time curves were
obtained. The upslope value of the line from the foot point
to the signal maximum was used for further calculations.
The myocardial perfusion reserve index was calculated by
division of the corrected upslope of the stress examination
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
by the corresponding segment’s corrected upslope value of
the rest examination. Microvascular obstruction was identified as described previously (9). All MR analyses were
performed in consensus by 3 independent experienced
radiologists (B.J.W., A.H., A.W.) (8) who were unaware of
study treatment, clinical, or laboratory data of study subjects.
Statistical analysis. Results are expressed in mean values ⫾
SD, or median (range) as indicated. Parametrical tests
included paired t testing; categoric variables were assessed
using chi-square or Fisher exact test where appropriate. For
perfusion analysis, non-parametrical tests (Mann-Whitney
U test, Wilcoxon test) were used. A level of p ⬍ 0.05 was
taken to indicate statistical significance (SPSS release 13.0,
SPSS Inc., Chicago, Illinois).
RESULTS
The baseline characteristics were comparable in both groups
(Table 1). All patients suffered from a partial or complete
proximal occlusion (Thrombolysis In Myocardial Infarction
[TIMI] flow grade 0/1) of at least 1 coronary artery
resulting in extensive MIs. Percutaneous coronary intervention was successfully achieved in all patients resulting in
TIMI flow grade 2/3. The time from PCI to onset of stem
cell mobilization or placebo treatment was comparable.
There was no significant difference between the groups
regarding cardiovascular risk factors, thrombolysis before
PCI, or use of glycoprotein IIb/IIIa antagonists. We used
stent sizes from 2.5 mm to 3.5 mm. Mean stent diameters,
or number of implanted stents, were comparable (data not
shown).
Stem cell mobilization and laboratory parameters. Granulocyte colony-stimulating factor treatment resulted in a
significant mobilization of different stem cell populations
(Table 2, Fig. 2). Granulocyte colony-stimulating factor
resulted in a transiently 4-fold increase of leukocytes when
compared with placebo. The number of endothelial precursors increased 23- to 29-fold after G-CSF but not after
placebo treatment. Both groups presented moderately elevated inflammatory parameters before treatment, which
decreased during treatment (Table 2).
Safety of G-CSF administration. Granulocyte colonystimulating factor was well tolerated in most patients.
During the application of either G-CSF or placebo, no
MACE were observed. In 2 patients, G-CSF was discontinued at day 3 because of occurrence of bone pain and
pericardial effusion, respectively. The patient presenting
moderate to severe bone pain during application of G-CSF
suffered from a common cold. Pericardial effusion after MI
was also documented in 1 case treated with placebo.
During the observational period, several MACE were
documented in both treatment groups. One patient of the
G-CSF group presented with a recurrent MI resulting from
in-stent thrombosis 14 days after initial stent procedure,
although the patient actually had taken aspirin and clopidogrel. Initially, this patient had received 3 stents with a
cumulative length of approximately 45 mm. The occlusion
Table 1. Baseline Characteristics of the Study Population
Therapy Stratum
Patients (n)
Male sex (%)
Age (yrs)
Cardiovascular risk factors
Hypertension (%)
Diabetes mellitus (%)
Smoking (%)
Hypercholesterolaemia (%)
Percutaneous coronary intervention
Infarct related artery (LAD/CX/RCA)
Peak creatinin kinase (U/l)
Angina to PCI, h (range)
Use of glycoprotein IIb/IIIa inhibitor (%)
Thrombolysis prior to PCI (%)
Onset of stem cell mobilization/placebo
treatment after PCI, h (range)
Time from PCI to baseline MRI (days)
Medication at 3-month follow-up
Aspirin (%)
Clopidogrel (%)
Beta-blocker (%)
ACE inhibitors (%)
Diuretics (%)
1715
G-CSF
Placebo
p Value
23
87
60 ⫾ 11
21
91
57 ⫾ 11
NS
NS
74
22
65
57
71
24
52
43
NS
NS
NS
NS
11/3/9
2,355 ⫾ 1,803
32 ⫾ 45
(6–160)
65
17
31 ⫾ 24
(2–107)
8.2 ⫾ 2.6
12/3/6
3,021 ⫾ 2,302
51 ⫾ 53
(7–168)
43
14
39 ⫾ 28
(13–135)
7.7 ⫾ 2.9
NS
NS
NS
95
52
100
100
42
100
29
100
100
29
NS
NS
NS
NS
NS
NS
NS
NS
NS
Values represent mean ⫾ SD.
ACE ⫽ angiotensin-converting enzyme; CX ⫽ circumflex artery; G-CSF ⫽ granulocyte colony-stimulating factor; LAD ⫽
left anterior descending artery; MRI ⫽ magnetic resonance imaging; PCI ⫽ percutaneous coronary intervention; RCA ⫽ right
coronary artery.
1716
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Table 2. Changes of Laboratory Parameters Including Stem
Cell Populations
WBC (g/l)
Baseline
Day 5
4 weeks
12 weeks
CD34⫹/CD133⫹ (cells/␮l)
Baseline
Day 5
CD34⫹/CD31⫹ (cells/␮l)
Baseline
Day 5
CD34⫹/c-kit⫹ (cells/␮l)
Baseline
Day 5
CRP (mg/dl)
Baseline
Day 5
4 weeks
12 weeks
IL-6 (ng/ml)
Baseline
Day 5
4 weeks
12 weeks
G-CSF
Placebo
p Value
11.2 ⫾ 4
42.9 ⫾ 25.7*
6.8 ⫾ 1.9
7.1 ⫾ 1.8
9.2 ⫾ 2.4
7.7 ⫾ 2.3
6.4 ⫾ 2.0
7.4 ⫾ 2.4
0.038
⬍0.001
NS
NS
1.6 ⫾ 0.9
46.1 ⫾ 33†
1.8 ⫾ 1.0
2.2 ⫾ 1.4
NS
⬍0.001
2.0 ⫾ 0.9
46.4 ⫾ 32.9‡
2.2 ⫾ 1.0
2.2 ⫾ 1.4
NS
⬍0.001
2.5 ⫾ 1.4
41.2 ⫾ 26.8§
2.2 ⫾ 1.2
3.3 ⫾ 3.3
NS
⬍0.001
5.8 ⫾ 6.4
4.1 ⫾ 2
0.5 ⫾ 0.2
0.5 ⫾ 0.3
8.4 ⫾ 8.4
2.5 ⫾ 3.3
0.6 ⫾ 0.1
0.5 ⫾ 0.1
NS
0.062
NS
NS
64 ⫾ 67
18 ⫾ 13
5.2 ⫾ 2.3
5.2 ⫾ 3.8
50 ⫾ 70
12 ⫾ 12
4.9 ⫾ 2.9
4.6 ⫾ 2.6
NS
NS
NS
NS
Values represent mean ⫾ SD. *p ⬍ 0.001 between baseline and day 5; †p ⫽ 0.002
between baseline and day 5; ‡p ⫽ 0.003 between baseline and day 5; §p ⫽ 0.008
between baseline and day 5.
CRP ⫽ C-reactive protein; IL ⫽ interleukin; WBC ⫽ white blood count; other
abbreviations as in Table 1.
could be successfully re-opened. One patient of the G-CSF
treatment group died 3 weeks after the 6-month angiography; autopsy examination was refused by the relatives. Two
patients of the placebo group, but none of the G-CSF
group, received coronary bypass grafts. Four of 19 patients
(21%) from the G-CSF group had an ISR of the infarctrelated artery and required target lesion revascularization
(TLR). Six of 21 patients (29%, p ⫽ 0.721 when compared
with G-CSF group) from the placebo group presented with
ISR, and subsequent TLR was performed (Table 3). Four
patients from the G-CSF group refused to undergo
follow-up angiography. None of them suffered from recurrent angina in the long term. When compared with placebo,
neither cumulative incidences of MACE nor occurrence of
restenosis differed significantly from the G-CSF group.
One patient of the G-CSF group presented stool abnormalities and malaise at 3 months of follow-up. We found a
colon carcinoma that was treated using chemotherapy. The
comparison of the observed serious adverse events demonstrated no significant difference of clinical events in both
treatment groups (Table 3).
Myocardial function and perfusion assessed by MRI. Baseline, 1-month, and 3-month follow-up MRI were performed 7.9 ⫾ 2.7 days, 41.0 ⫾ 17.2 days, and 108.6 ⫾ 23.4
days after PCI, respectively. Magnetic resonance imaging of
37 patients who completed 3-month follow-up (G-CSF
group: n ⫽ 19, placebo: n ⫽ 18) demonstrated that left
ventricular EF was comparable at baseline (1 week after
PCI) in both groups. Ejection fraction improved by 6.2 ⫾
9.0% from baseline to 3 months of follow-up in G-CSFtreated patients and by 5.3 ⫾ 9.8% after placebo treatment
(p ⫽ 0.77) (Fig. 3, Table 4). Other global and regional
myocardial function parameters are shown in Table 4.
Infarct volumes were comparable and decreased significantly
from baseline to 3 months of follow-up in the G-CSF
treatment group. The infarct volume in the placebo group
was reduced by trend. Microvascular obstruction was observed
in 33% (G-CSF) and 29% (placebo) of cases (p ⫽ 1.0).
Granulocyte colony-stimulating factor resulted in a significantly increased resting perfusion in the area of infarction at baseline, and at 1 month of follow-up (Table 5).
Resting perfusion at 3 months was slightly increased in
G-CSF-treated patients. Adenosine-induced hyperemia resulted in a significant increase of perfusion in both G-CSFand placebo-treated subjects in infarct as well as in remote
areas. The myocardial perfusion reserve index was slightly
increased in infarct and remote areas in placebo-treated
patients compared with the G-CSF group.
DISCUSSION
The present findings demonstrate, for the first time, safety
and feasibility of G-CSF treatment in patients suffering
from late revascularized STEMI in a prospective, randomized, placebo-controlled analysis. Granulocyte colonystimulating factor was not superior to placebo regarding
improvement of global as well as regional myocardial
function. As a secondary finding, the trial demonstrates
significant increase of myocardial perfusion in the short
term. Several stem-cell populations, which are considered to
improve myocardial regeneration or neovascularization, are
significantly mobilized by G-CSF. This cytokine was generally well tolerated, with no significantly higher rate of ISR
or MACE when compared with placebo.
The novel therapeutic concept of stem-cell mobilization
using G-CSF in MI was made uncertain by Kang et al. (5),
who described ISR to be a major adverse event. In contrast
with this report, the present study and 4 recent publications
of clinical trials report rare occurrences of ISR or TLR
(3,4,10,11). In addition, intravascular ultrasound in humans
demonstrated no increased neointima formation in G-CSFtreated subjects (12). The lower rate of ISR in our study
patients may be influenced by the time point of G-CSF
administration after establishment of complete revascularization. Complete stem-cell mobilization in the MAGIC
cell trial (effects of intracoronary infusion of peripheral
blood stem-cells mobilized with G-CSF on left ventricular
systolic function and restenosis after coronary stenting in
MI) was achieved 4 days before elective PCI of the
infarct-related artery (5) at the time of highest leukocyte
count, which may have resulted in proangiogenic and
proinflammatory processes within the culprit lesion. Hill et
al. (13) recently reported another clinical study of 16
patients suffering from reproducible myocardial ischemia in
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Engelmann et al.
G-CSF in Myocardial Infarction
1717
Figure 2. Flow cytometry analysis of mobilized stem cells at baseline and 5 days after initiation of granulocyte colony-stimulating factor (G-CSF). CD34⫹
stem cell populations from peripheral blood measured by flow cytometry. (A) Quantitative measurements of CD34⫹/CD31⫹, CD34⫹/133⫹, and
CD34⫹/CD117⫹ cells in placebo and G-CSF-treated patients before and at day 5 of treatment. *p ⬍ 0.001; **p ⬍ 0.001; ***p ⬍ 0.001 when compared
with placebo at day 5. (B) The R4 population demonstrates increases of precursor cells in a representative patient before and at day 5 of treatment with
G-CSF.
whom coronary revascularization had not been performed
before G-CSF administration. In this trial, 2 cases of MI
were observed, and 1 patient died 17 days after G-CSF
treatment (13). In contrast with our trial and the more
recent studies (12,14), Hill et al. (13) did not administer
clopidogrel or glycoprotein IIb/IIIa antagonists during
G-CSF treatment.
However, we documented several adverse events in
G-CSF-treated subjects. One patient, who suffered from
MI owing to an in-stent thrombosis, had received 3 stents
with a total length of 45 mm into the left anterior descending coronary artery followed by clopidogrel and tirofiban
during the initial procedure. In-stent restenosis and thrombosis are reported to be associated with an increased lesion
and stent length (15). One patient of the G-CSF group died
3 weeks after the 6-month follow-up angiography, which
had demonstrated a good stent result and a normal myocardial function. Death may have resulted from sudden
cardiac death, although ventricular arrhythmias have never
been documented in his medical history. One case of colon
carcinoma was observed in a patient from the G-CSF group
after 3 months. The occurrence of a formerly undiagnosed
intestinal tumor may be coincidental, because human bone
marrow donors, who receive G-CSF, do not show significant increase of malignancies (16). Granulocyte colonystimulating factor was shown to have no effect on cancer cell
proliferation in mice but promoted tumor growth via
enhanced angiogenesis (17).
In the present study, left ventricular function improved
significantly in both treatment groups, but G-CSF was not
superior to placebo. Most animal studies reported improved
hemodynamic and myocardial function when G-CSF was
administered before or after establishment of infarction
(2,18). One mechanism of repair, by which G-CSF may
improve cardiac function after MI, is considered to be the
mobilization of bone marrow– derived stem cells homing
into the damaged tissue area, where they induce neovascularization (18,19). Homing of the stem cells can be im-
1718
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Table 3. Safety Characteristics of the Study Population
Therapy Stratum
G-CSF
(n ⴝ 23)
Major adverse cardiac events
Death
(Re) myocardial infarction
Acute coronary syndrome
Coronary artery bypass grafting
Target lesion revascularisation because of in-stent
restenosis*
De novo lesion requiring PCI*
Spontaneously reported adverse events
Bone/muscle pain
Pneumonia
Pericardial effusion
Local inflammation at arterial puncture site
Colon carcinoma
Sustained CRP elevation of unknown origin
Cumulative spontaneously reported adverse events
Placebo
(n ⴝ 21)
p Value
1 (4)
1 (4)
0
0
0
0
0
2 (10)
NS
NS
4 (21)*
4 (21)*
6 (29)*
2 (10)*
NS
NS
1 (4)
2 (9)
1 (4)
1 (4)
1 (4)
0
7 (30)
0
0
1 (5)
0
0
1 (5)
3 (14)
NS
NS
NS
NS
NS
NS
NS
NS
Values are n (%). *19 of 23 patients from the G-CSF group and 21 of 21 patients from the placebo group underwent follow-up
coronary angiography.
Abbreviations as in Tables 1 and 2.
proved by the chemokine stromal cell– derived factor 1
(SDF-1), which is intrinsically produced by the myocardium after MI. As recently shown, the intramyocardial
delivery of SDF-1 combined with G-CSF increases homing
of c-kit⫹ stem cells (20). On the other hand, direct
antiapoptotic effects of G-CSF via activation of the Jak/Stat
pathway may contribute to an improved survival of cardiomyocytes preventing left ventricular remodeling after MI
(21). Recently, the G-CSF receptor was shown to be
up-regulated shortly after MI, indicating a sensitization of
the heart to direct influences of this specific cytokine (22).
The application of G-CSF in humans after MI results in
a more heterogeneous pattern of effects. The FIRSTLINEAMI (Front-Integrated Revascularization and Stem Cell
Liberation in Evolving Acute Myocardial Infarction by
Granulocyte Colony-Stimulating Factor) trial demonstrated
in a randomized and controlled, but not blinded, study a
beneficial effect of G-CSF on MI (4,14). Treatment of 25
patients with G-CSF first given 85 ⫾ 30 min after immediately performed PCI resulted in a significant improvement
of EF of 8% at 12 months, while EF of non-placebo
controlled subjects decreased by 5%. Another study ob-
Figure 3. Change of global ejection fraction assessed by magnetic resonance imaging from baseline (1 week after percutaneous coronary intervention) to
3 months of follow-up. G-CSF ⫽ granulocyte colony-stimulating factor; PCI ⫽ percutaneous coronary intervention. Bold squares ⫽ mean ⫾ SD.
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
1719
Table 4. Myocardial Function Parameters Assessed by Magnetic Resonance Imaging
Therapy Stratum
Ejection fraction
Baseline (1 week after PCI) (%)
3 months (%)
p value†
⌬ ejection fraction (% points)
LVEDV
Baseline (1 week after PCI) (ml)
3 months (ml)
p value†
⌬ LVEDV (ml)
LVESV
Baseline (1 week after PCI) (ml)
3 months (ml)
p value†
⌬ LVESV (ml)
Stroke volume
Baseline (1 week after PCI) (ml)
3 months (ml)
p value†
⌬ stroke volume (ml)
Systolic myocardial thickening of infarct area
Baseline (1 week after PCI) (mm)
3 months (mm)
p value†
⌬ SMTinfarct area (mm)
Systolic myocardial thickening of infarct border zone
Baseline (1 week after PCI) (mm)
3 months (mm)
p value†
⌬ SMTborderzone (mm)
Systolic myocardial thickening of remote area
Baseline (1 week after PCI) (mm)
3 months (mm)
p value†
⌬ SMTremote area (mm)
Infarct volume
Baseline (1 week after PCI) (ml)
3 months (ml)
p value†
⌬ infarct volume (ml)
G-CSF
(n ⴝ 19)
Placebo
(n ⴝ 18)
41 ⫾ 12
47 ⫾ 12
0.007
6.2 ⫾ 9.0
44 ⫾ 9
49.5 ⫾ 12
0.035
5.3 ⫾ 9.8
NS
NS
140 ⫾ 44
145 ⫾ 61
NS
2 ⫾ 42
149 ⫾ 32
162 ⫾ 52
NS
13 ⫾ 46
NS
NS
83 ⫾ 38
79 ⫾ 46
NS
⫺5 ⫾ 26
87 ⫾ 21
85 ⫾ 44
NS
⫺2 ⫾ 35
NS
NS
57 ⫾ 21
65 ⫾ 21
NS
6 ⫾ 24
63 ⫾ 20
77 ⫾ 22
0.041
15 ⫾ 28
NS
NS
2.7 ⫾ 2.3
3.1 ⫾ 1.5
NS
0.5 ⫾ 2.0
3.1 ⫾ 2.5
3.6 ⫾ 1.7
NS
0.9 ⫾ 2.1
2.9 ⫾ 1.6
3.5 ⫾ 2.6
NS
0.4 ⫾ 2.3
3.0 ⫾ 1.9
3.2 ⫾ 2.0
NS
0.5 ⫾ 2.2
6.0 ⫾ 1.8
5.6 ⫾ 2.5
NS
0.6 ⫾ 2.5
6.6 ⫾ 3.1
6.5 ⫾ 2.2
NS
0.6 ⫾ 2.3
30 ⫾ 11
25 ⫾ 11
0.015
⫺5 ⫾ 6
35 ⫾ 20
31 ⫾ 18
0.08
⫺5 ⫾ 11
p Value*
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
NS
Values represent mean ⫾ SD. *Between treatment groups; †between baseline (1 week) and 3-month follow-up magnetic
resonance imaging.
LVEDV ⫽ left ventricular end-diastolic volume; LVESV ⫽ left ventricular end-systolic volume; SMT ⫽ systolic myocardial
thickening; other abbreviations as in Table 1.
served a tendency of improvement of EF and left ventricular
end-diastolic volume in G-CSF-treated subjects (11). In
this study, patients with late presentation were not submitted to primary PCI. Granulocyte colony-stimulating factor/
placebo was initiated at a lower dose and more rapid than in
the present study. Recently, in a prospective, nonrandomized, open-label study (3), 14 patients were treated
with G-CSF 48 h after PCI for 7 ⫾ 1 days. The EF
increased by 8% in the G-CSF group, compared with 3% in
the control group. In contrast with the present investigation,
these studies may be biased either by the absence of
double-blind placebo treatment or the selection of control
subjects from patients, who refused G-CSF treatment.
More recently, a double-blind, randomized, placebo-
controlled trial investigated the use of G-CSF versus placebo after immediately performed PCI in patients suffering
from acute MI (n ⫽ 87) (10). The time point of G-CSF
initiation in this trial was comparable to our study, but PCI
was performed earlier. The systolic wall thickening improved by 17% in the infarct area of both G-CSF- and
placebo-treated patients. Both groups demonstrated improvement of EF by 8%, but G-CSF was not superior to
placebo.
In contrast to the variety of recently published clinical
trials, the present study focused on the effects of G-CSF in
patients suffering from subacute MI who were admitted late
for PCI. The outcome of patients suffering from subacute
STEMI is considered to be serious; patients are threatened
1720
Engelmann et al.
G-CSF in Myocardial Infarction
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
Table 5. Myocardial Perfusion Assessed by Magnetic Resonance Imaging
Therapy Stratum
G-CSF
(n ⴝ 17)
Perfusion at rest (Upslope) [signal intensity]
Infarct area
Baseline (1 week after PCI)
1.1 (0.3–2.0)
1 month
1.1 (0.4–5.6)
3 months
1.0 (0.4–2.1)
p value†
NS
Remote area
Baseline (1 week after PCI)
1.2 (0.2–3.3)
1 month
1.1 (0.4–7.9)
3 months
1.0 (0.4–2.5)
p value†
NS
Perfusion with adenosine stress (Upslope) [signal intensity]
Infarct area
1 month
2.3 (0.3–8.0)‡
3 months
1.9 (0.9–3.0)‡
p value
NS
Remote area
1 month
2.8 (0.4–8.9)‡
3 months
1.9 (0.9–4.4)‡
p value
NS
MPRI (Upslopestress/Upsloperest)
Infarct area
1 month
1.5 (0.5–5.8)
3 months
1.7 (0.6–2.6)
p value
NS
Remote area
1 month
1.7 (0.9–3.1)
3 months
1.7 (0.4–3.6)
p value
NS
Placebo
(n ⴝ 17)
p Value*
0.7 (0.3–1.5)
0.7 (0.1–1.9)
0.8 (0.1–2.4)
NS
0.048
0.016
NS
0.9 (0.2–2.2)
0.7 (0.2–1.6)
0.9 (0.1–1.9)
NS
NS
NS
NS
1.4 (0.3–3.4)§
1.5 (0.1–3.9)§
NS
NS
NS
1.6 (0.6–4.8)§
1.8 (0.1–3.6)§
NS
NS
NS
2.3 (0.7–13)
2.0 (0.2–29)
NS
NS
NS
2.1 (0.8–7.3)
2.2 (0.1–32)
NS
NS
NS
Values represent median (range). *Between treatment groups; †between baseline and 1 month F/U; ‡p ⬍ 0.005 when compared
to Upslope at rest; §p ⬍ 0.005 when compared to upslope at rest.
MPRI ⫽ myocardial perfusion reserve index; other abbreviations as in Table 1.
by progressive myocardial dysfunction resulting from the
long time interval between onset of infarction to revascularization (1). The initial EFs were markedly reduced in
both study groups reflecting an extended myocardial damage in those patients. Most other trials included patients
presenting less severe myocardial dysfunction (4,23). Microvascular obstruction, which may have contributed to the
observed myocardial dysfunction, occurred in both treatment groups within a minority of patients in equal measure.
The time point of G-CSF initiation after successful PCI
may be an important factor, because G-CSF treatment was
started very rapidly in the FIRSTLINE-AMI trial. In our
study, G-CSF and placebo treatment were initiated at an
average of 31 ⫾ 24 h after successful revascularization. In a
larger trial of 114 subjects (G-CSF n ⫽ 56) (23), patients
received G-CSF 5 days after immediately performed PCI.
The very late treatment resulted in a small decrease of
infarct size in both groups (G-CSF ⫺6.2 ⫾ 9.1% vs. ⫺4.9
⫾ 8.9% in placebo control subjects, p ⫽ 0.56) and no
substantial increase of EF in both groups. This result
supports the hypothesis that late administration after PCI
may diminish the potential benefit of G-CSF as demonstrated in animal studies (21).
In the current trial, we demonstrate an increased myocardial perfusion at rest in G-CSF-treated patients within 1
month. The reason for elevated perfusion assessed by
baseline MR may result from the late time point of
examination. Baseline MR was performed an average of 8
days after PCI. At that point, the 5-day course of G-CSF
was already administered, and early effects of G-CSF on
myocardial perfusion may have occurred. The beneficial
effect of G-CSF on myocardial perfusion has not been
established in humans. However, increased cardiac blood
flow and metabolism after G-CSF treatment was shown in
a baboon model of infarction (24). Mechanisms that promote enhanced myocardial perfusion include significant
release of endothelial progenitor cells into circulation, as
observed in our study. In addition, paracrine angiogenic
factors, such as vascular endothelial growth factor, are
released by neutrophils after administration of G-CSF and
increase neovascularization in ischemic tissue (19). Enhanced neovascularization may serve as one important
mechanism facilitating reduction of infarct size after MI.
The reduction of infarct size was significant in the G-CSF
group, but less so in the control group (p ⫽ 0.15 vs. p ⫽
0.08). Interestingly, the absolute infarct size was smaller in
JACC Vol. 48, No. 8, 2006
October 17, 2006:1712–21
G-CSF-treated patients. We recently demonstrated in a
murine model of infarction that G-CSF results in a reduced
infarct size and an enhanced arteriogenesis in the periinfarct area mediated by an increased expression of the
intracellular adhesion molecule-1 on endothelial cells (25).
In summary, G-CSF treatment appears to be safe in the
majority of subjects suffering from MI, when successful PCI
was performed. The occurrence of 1 death and 1 MI in the
G-CSF group necessitates a careful patient monitoring
during further studies. Granulocyte colony-stimulating factor was not superior to placebo regarding improvement of
myocardial function in patients with subacute infarctions in
whom delayed PCI was performed. Granulocyte colonystimulating factor resulted in a significant increase of myocardial perfusion within 1 month after PCI. Due to its phase
II character, the present study is limited by the relatively low
number of patients. We conclude that further research
should focus on immediate administration of G-CSF in
early revascularized MI and on larger multicenter trials
investigating clinical outcome.
Acknowledgments
The authors thank Dr. H. Diem and Dr. M. Adam,
Hematology Laboratory, Institute of Clinical Chemistry,
Klinikum Grosshadern, Munich, for their excellent support.
Reprint requests and correspondence: Prof. Dr. med.
Wolfgang-M. Franz, Ludwig Maximilians University, Medical
Clinic I—Department of Cardiology Klinikum Grosshadern,
Marchioninistr. 15, D-81377 Munich, Germany. E-mail:
[email protected].
REFERENCES
1. Cannon CP, Gibson CM, Lambrew CT, et al. Relationship of
symptom-onset-to-balloon time and door-to-balloon time with mortality in patients undergoing angioplasty for acute myocardial infarction. JAMA 2000;283:2941–7.
2. Orlic D, Kajstura J, Chimenti S, et al. Mobilized bone marrow cells
repair the infarcted heart, improving function and survival. Proc Natl
Acad Sci U S A 2001;98:10344 –9.
3. Kuethe F, Figulla HR, Herzau M, et al. Treatment with granulocyte
colony-stimulating factor for mobilization of bone marrow cells in
patients with acute myocardial infarction. Am Heart J 2005;150:115.
4. Ince H, Petzsch M, Kleine HD, et al. Prevention of left ventricular
remodeling with granulocyte colony-stimulating factor after acute
myocardial infarction: final 1-year results of the Front-Integrated
Revascularization and Stem Cell Liberation in Evolving Acute Myocardial Infarction by Granulocyte Colony-Stimulating Factor
(FIRSTLINE-AMI) Trial. Circulation 2005;112:I73– 80.
5. Kang HJ, Kim HS, Zhang SY, et al. Effects of intracoronary infusion
of peripheral blood stem-cells mobilised with granulocyte-colony
stimulating factor on left ventricular systolic function and restenosis
after coronary stenting in myocardial infarction: the MAGIC cell
randomised clinical trial. Lancet 2004;363:751– 6.
6. Bolli R, Triana JF, Jeroudi MO. Prolonged impairment of coronary
vasodilation after reversible ischemia. Evidence for microvascular
“stunning.” Circ Res 1990;67:332– 43.
Engelmann et al.
G-CSF in Myocardial Infarction
1721
7. Wintersperger BJ, Reeder SB, Nikolaou K, et al. Cardiac CINE MR
imaging with a 32-channel cardiac coil and parallel imaging: impact of
acceleration factors on image quality and volumetric accuracy. J Magn
Reson Imaging 2006;23:222–7.
8. Huber A, Bauner K, Wintersperger BJ, et al. Value of phase-sensitive
inversion recovery (PSIR) for detection of myocardial infarction.
Radiology 2005;237:854 – 60.
9. Gerber BL, Garot J, Bluemke DA, Wu KC, Lima JA. Accuracy of
contrast-enhanced magnetic resonance imaging in predicting improvement of regional myocardial function in patients after acute myocardial
infarction. Circulation 2002;106:1083–9.
10. Ripa RS, Jorgensen E, Wang Y, et al. Stem cell mobilization induced
by subcutaneous granulocyte-colony stimulating factor to improve
cardiac regeneration after acute ST-elevation myocardial infarction.
Result of the double-blind, randomized, placebo-controlled Stem
Cells in Myocardial Infarction (STEMMI) Trial. Circulation 2006;
113:1983–92.
11. Valgimigli M, Rigolin GM, Cittanti C, et al. Use of granulocytecolony stimulating factor during acute myocardial infarction to enhance bone marrow stem cell mobilization in humans: clinical and
angiographic safety profile. Eur Heart J 2005;26:1838 – 45.
12. Jorgensen E, Ripa RS, Helqvist S, et al. In-stent neo-intimal hyperplasia after stem cell mobilization by granulocyte-colony stimulating
factor. Preliminary intracoronary ultrasound results from a doubleblind randomized placebo-controlled study of patients treated with
percutaneous coronary intervention for ST-elevation myocardial infarction (STEMMI Trial). Int J Cardiol 2006;111:174 –7.
13. Hill JM, Syed MA, Arai AE, et al. Outcomes and risks of granulocyte
colony-stimulating factor in patients with coronary artery disease.
J Am Coll Cardiol 2005;46:1643– 8.
14. Ince H, Petzsch M, Kleine HD, et al. Preservation from left ventricular remodeling by Front-Integrated Revascularization and Stem Cell
Liberation in Evolving Acute Myocardial Infarction by Use of
Granulocyte-Colony-Stimulating Factor (FIRSTLINE-AMI). Circulation 2005;112:3097–106.
15. Kastrati A, Elezi S, Dirschinger J, et al. Influence of lesion length on
restenosis after coronary stent placement. Am J Cardiol 1999;83:1617–
22.
16. Cavallaro AM, Lilleby K, Majolino I, et al. Three to six year follow-up
of normal donors who received recombinant human granulocyte
colony-stimulating factor. Bone Marrow Transplant 2000;25:85–9.
17. Natori T, Sata M, Washida M, et al. G-CSF stimulates angiogenesis
and promotes tumor growth: potential contribution of bone marrowderived endothelial progenitor cells. Biochem Biophys Res Commun
2002;297:1058 – 61.
18. Ohtsuka M, Takano H, Zou Y, et al. Cytokine therapy prevents left
ventricular remodeling and dysfunction after myocardial infarction
through neovascularization. FASEB J 2004;18:851–3.
19. Ohki Y, Heissig B, Sato Y, et al. Granulocyte colony-stimulating
factor promotes neovascularization by releasing vascular endothelial
growth factor from neutrophils. FASEB J 2005;19:2005–7.
20. Askari AT, Unzek S, Popovic ZB, et al. Effect of stromal-cell-derived
factor 1 on stem-cell homing and tissue regeneration in ischaemic
cardiomyopathy. Lancet 2003;362:697–703.
21. Harada M, Qin Y, Takano H, et al. G-CSF prevents cardiac
remodeling after myocardial infarction by activating the Jak-Stat
pathway in cardiomyocytes. Nat Med 2005;11:305–11.
22. Kuhlmann MT, Kirchhof P, Klocke R, et al. G-CSF/SCF reduces
inducible arrhythmias in the infarcted heart potentially via increased
connexin43 expression and arteriogenesis. J Exp Med 2006;203:87–97.
23. Zohlnhofer D, Ott I, Mehilli J, et al. Stem cell mobilization by
granulocyte colony-stimulating factor in patients with acute myocardial infarction: a randomized controlled trial. JAMA 2006;295:1003–
10.
24. Norol F, Merlet P, Isnard R, et al. Influence of mobilized stem cells on
myocardial infarct repair in a nonhuman primate model. Blood
2003;102:4361– 8.
25. Deindl E, Zaruba MM, Brunner S, et al. G-CSF administration after
myocardial infarction in mice attenuates late ischemic cardiomyopathy
by enhanced arteriogenesis. FASEB J 2006;20:956 – 8.